Wireless entity sensing

Abstract

A wireless vision system may include a radio transceiver and a plurality of antennas including antennas positioned in at least two spatially separated banks, the radio transceiver configured to transmit and receive non-visual frequency photons. A wireless vision system may include a controller in communication with the radio transceiver and the plurality of antennas, the controller configured to: select antenna pairs from among the plurality of antennas, transmit configured messages between selected antennas of the wireless communication device over one or more nonvisual frequency channels, obtain channel state information that characterizes propagation of the non-visual frequency photons through a specified area, process the channel state information to infer one or more entities within the specified area; and generate a view that is a visualization of the at least one of the one or more entities in the specified area based on the channel state information.

Description

Attorney Docket No. CURV-0003-WQWIRELESS ENTITY SENSING CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims the benefit of U.S. Provisional Patent App. No. 63 / 755,152, filed on 6 FEB 2025, and entitled “WIRELESS ENTITY SENSING” (CURV-0001-P01).

[0002] The present application claims the benefit of U.S. Provisional Patent App. No. 63 / 760,493, filed on 19 FEB 2025, and entitled “WIRELESS ENTITY SENSING” (CURV-0002-P01).

[0003] Each of the foregoing applications is incorporated herein by reference in the entirety for all purposes.BACKGROUND

[0004] Previously known monitoring systems, including automated monitoring systems such as camera based monitoring and / or security systems and / or motion detection systems suffer from a number of drawbacks. For example, camera based monitoring systems are limited to the constraints of visual frequency photons, which include limitations such as an inability to perceive entities in the monitored space that are blocked by visually opaque objects, and / or that are not visible due to lighting limitations, such as areas that are dark and / or where it is desired that the area remain dark. Certain camera based monitoring systems may be based on photons outside of the visual frequency portion of the electromagnetic (EM) spectrum, such as infrared photons, but are still limited to visualizing objects that emit (e.g., thermal emissions) or reflect such photons off the surface of objects, and / or that are obscured by objects emitting such photons even if the objects are not opaque to the frequency being monitored (e.g., thermal photons emitted by an object obscure objects behind it, even if the object is not opaque to the infrared portion of the spectrum). On the other end of the spectrum, for example systems capable of monitoring the ultraviolet (UV) portion of the EM spectrum, tend to be even more limited due to object opacity (e.g., glass is typically opaque to UV photons). While UV systems may allow for illumination of a target area in a manner that will not directly disturb the area visually, such systems nevertheless cannot see through UV opaque objects, and nevertheless tend to disturb the area visually through secondary emissions such as via fluorescence. Further, it can be undesirable to flood an area with significant UV radiation.

[0005] Further, previously known systems that utilize visual or near- visual photons may raise privacy concerns, for example such systems are problematic for monitoring private areas such as bathrooms and / or living spaces. In addition to privacy concerns, visual monitoring of spaces can also disturb users more generally, creating a greater perception of being watched and / or monitored, which can be even more disturbing to monitored people when the entity performing the monitoring has a power dynamic with monitored people in the target space, such as when the monitoring isAttorney Docket No. CURV-0003-WQperformed by an employer, a government agency, management of a care facility, and / or a school administrator. Such disturbance can facilitate a negative experience for people interacting with the monitored space, which may reduce benefits to people overall in utilizing services associated with the space. Additionally or alternatively, such disturbance can degrade perceptions of the monitoring entity with attendant negative consequences such as developing a negative perception of employers, corporations, or government agencies, with attendant negative consequences such as reducing the benefits of government services to people that could otherwise benefit from those services, reducing employee morale, and / or reducing employee productivity.

[0006] Still further, previously known systems that utilize visual or near-visual photons may increase liability and / or compliance concerns, and / or compliance costs, for the monitoring agencies. For example, visual system may capture personally identifiable information, and / or facilitate the collection of personally identifiable information, even unintentionally, through the combination of the captured visual data with other information that may be present in related data sets where the visual information may be stored. Where personally identifiable information is captured and / or created, various responsibilities may be imputed to the monitoring entity to secure and / or delete the data, and / or to notify related persons that the data has been collected. Further, systems that capture visual data, including visual data that captures and / or creates personally identifiable information increases the costs to train and monitor relevant employees, as employees that fail to conform to usage standards and / or misuse the data (whether intentionally or unintentionally) create liability hazards and / or compliance risks for the monitoring entity.SUMMARY

[0007] A wireless vision system transmits configured messages over non-visual frequency channels using a radio transceiver and antennas arranged in at least two spatially separated banks, acquires channel state information for those messages, and processes the information to infer entities and generate a visualization of a specified area. These elements collectively support detecting entities without illumination by visual or near-visual photons and operating in a qualia not inherently recognizable to humans, with optional conversion to human-interpretable visualizations. Selection of antenna pairs across fixed baselines, together with polarization separation within banks and cross-bank pairing, increases channel decorrelation and angular diversity to enable sensing through obstacles that impair visual systems and to provide robust entity presence and location determinations. Utilization of channel state information across multiple non-visual frequency ranges facilitates extraction of propagation features well-suited to identifying materials and volumes, thereby supporting detection of large metal objects indicative of potential threats or items of interest. Implementations employing independent driving circuits, solid-state components, and multi-bandAttorney Docket No. CURV-0003-WQoperation over standard radio and microwave spectrum leverage readily available power and network infrastructure, reducing installation cost and complexity while enabling compact devices housed within a single enclosure. By sequencing configured messages across antenna banks without requiring line of sight to all monitored regions, the system can monitor remote or lightly traveled areas. Further, because the system inherently does not rely on camera imagery, it supports monitoring of sensitive areas without producing human-interpretable qualia at all, or alternatively producing a visualization only upon detection of specified events, thereby reducing privacy and compliance risks. A given system may support some or all of the recited benefits, and / or may support other benefits not listed.

[0008] A wireless vision system utilizes a radio transceiver and antennas arranged in at least two spatially separated banks to transmit configured messages over non-visual frequency channels, acquire channel state information for those messages, and process the information to create an entity model comprising multiple coefficients and joints for each entity. Associated methods select antenna pairs across banks, obtain channel state information, and output the entity model for use in visualization operations including tracking, identifying, alerting, and displaying. These elements collectively support detecting entities without illumination by visual or near-visual photons and operating in a qualia not inherently recognizable to humans, with optional conversion tohuman-interpretable visualizations. Selection of antenna pairs across fixed baselines together with polarization separation and cross-bank pairing increases channel decorrelation and angular diversity, enabling sensing through obstacles that impair visual systems and supporting robust determinations of entity presence and location. Utilization of channel state information across non-visual frequencies facilitates extraction of propagation features well-suited to identifying materials and volumes, thereby supporting detection of large metal objects and other items of interest.Implementations employing independent driving circuits, solid-state components, and multi-band operation over standard radio and microwave spectrum leverage readily available power and network infrastructure, reducing installation cost and complexity while enabling compact devices housed within a single enclosure. Sequencing configured messages across antenna banks without requiring line of sight to all monitored regions allows monitoring of remote or lightly traveled areas, and operation without camera imagery enables monitoring of sensitive areas without producing human-interpretable qualia at all, or alternatively producing a visualization only upon detection of specified events, thereby reducing privacy and compliance risks. A given system may support some or all of the recited benefits, and / or may support other benefits not listed.

[0009] Detecting a metal entity in the target area provides immediate security and safety benefits by enabling early identification of objects with significant conductive or ferromagnetic content that mayAttorney Docket No. CURV-0003-WQindicate weapons, tools, or concealed containers. This detection supports proactive risk mitigation at entry ways and sensitive zones, facilitates targeted alerts and interventions without relying on camera imagery, and preserves privacy while maintaining situational awareness. It further allows the system to trigger condition-based visualizations only when a metal signature is present, reducing compliance burdens and data handling overhead.

[0010] Determining a relationship between the detected metal entity and other detected entities provides context that improves decision quality and operational response. For example, establishing proximity, carriage (e.g., on-person or in-bag), or placement relative to furnishings or structures allows prioritization of alerts, differentiation between benign objects and potential threats, and more accurate assessment of activities (e.g., carrying, setting down, or transferring an item). This relational information supports smarter workflows such as directing personnel to specific locations, logging events for audits, and tailoring notifications to reduce false positives while maintaining privacy by avoiding visual imaging. A given system may support some or all of the recited benefits, and / or may support other benefits not listed.

[0011] A wireless vision system operates over non-visual frequency channels to transmit configured messages, acquire propagation data such as channel state information, and convert that information into human-interpretable outputs without reliance on camera imagery. A controller processes the measurements into a joint model for each detected entity, specifying joint locations, orientations, and associated parameters. Using the joint model provides direct benefits, including precise posture and orientation determination, improved activity inference from joint-state changes over time, enhanced continuity tracking of individuals, and richer context for alerting and visualization while preserving privacy by avoiding continuous visual imaging and rendering only upon specified events.

[0012] Environment placement derived from joint-model parameters positions entities relative to structures and human-use furnishings, such as seating and work surfaces, improving situational understanding for safety, access control, and workflow decisions. Visualizations with annotations based on environment placement convey who is seated, leaning, or interacting with a surface, enabling targeted notifications and audit logs while reducing compliance burdens through limited generation of human-interpretable qualia. Implementations that leverage standard radio hardware and solid-state components within compact enclosures use readily available power and network infrastructure, reducing installation cost, complexity, and footprint, and enabling monitoring of remote or lightly traveled areas without requiring line of sight.

[0013] A given system may support some or all of the recited benefits, and / or may support other benefits not listed.Attorney Docket No. CURV-0003-WQBRIEF DESCRIPTION OF THE FIGURES

[0014] Fig. 1 is an example procedure for providing entity vision operations utilizing a wireless vision system schematically depicted.

[0015] Fig. 2 is an example calibration location setup depicted.

[0016] Fig. 3 is example and non-limiting entity vision operations schematically depicted.

[0017] Fig. 4 is an example main vision device (or wireless vision device) schematically depicted.

[0018] Fig. 5 shows a schematic view of one example of a representative embodiment used for illustration.

[0019] Fig. 6 shows a side view and a front view of the antenna enclosure, a secondary enclosure (e.g., a controller enclosure) where various controllers and / or circuits are housed to support operations of the wireless vision device, and heat fins that support heat rejection from the wireless vision device.

[0020] Fig. 7 is an example bottom view of a wireless vision device depicting the antenna housing and an access cover that can be removed to allow access into the housing.

[0021] Fig. 8 shows an example configuration where target location (and / or specified area) is a building entrance / exit equipped with a ceiling-mounted device, which is part of the wireless vision system.

[0022] Fig. 9 is an example and non-limiting arrangement of a system positioned at a target location 902 as schematically depicted.

[0023] Fig. 10 is an example vision output of wireless vision systems schematically depicted.

[0024] Fig. 11 is a visual view of a scene as schematically depicted on the left side, and a wireless view of the scene schematically depicted on the right side.

[0025] Fig. 12 is a wireless view of a scene schematically depicted with a number of detected entities in various postures as may be seen by a wireless vision system.

[0026] Fig. 13 is an example schematically depicted wireless vision system.

[0027] Fig. 14 is an example procedure for generating a visualization of entities in a specified area.

[0028] Fig. 15 is an example procedure for creating an entity model for entities within a specified area.

[0029] Fig. 16 is an example procedure for determining a presence, location, and / or relationship of a metal entity with another entity, within a specified area.

[0030] Fig. 17 is an example procedure for calibrating a wireless visions system for a location of interest.

[0031] Fig. 18 is an example procedure for determining joint model parameters of an entity in a specified area.Attorney Docket No. CURV-0003-WQ

[0032] Fig. 19 is an example depiction of channel state information.DETAILED DESCRIPTION

[0033] Without limitation to any other aspect of the present disclosure, aspects of the disclosure herein provide for systems and methods to detect entities in a space using wireless signals, without utilization of camera visual information. Embodiments herein allow for detection and labeling of entities (e.g., people, pets, furniture, or other objects of interest) and further allow for the determination of actions of interest being performed by entities where applicable. Embodiments herein are supported by a simple discrete hardware setup, optionally utilizing off-the-shelf components which may be customized to support operations set forth herein.

[0034] Embodiments herein support determining the posture and positioning of entities, allowing for determination of events, actions, intentions, or the like. Embodiments herein do not require the utilization of a camera, although in certain embodiments combination of wireless sensing with cameras may enhance certain capabilities for those embodiments. Embodiments herein support the utilization of a number of wireless frequencies, including those utilized in commercial applications such as WiFi, Bluetooth, millimeter wave, or the like. Embodiments herein utilize channel state information (CSI), and / or other similar information such as phase and / or amplitude information related to wireless signals derived from any source, to provide the state space information for entity determination and other sensing aspects herein.

[0035] Example embodiments herein can determine the presence and location of entities, the positioning and / or posture of those entities, detect the presence and location of large metal devices and the relationship of those devices with entities in the detection space. Example embodiments herein can track individual entities, such as people, in either an instance identification (e.g., confirming the same entity as it moves through the detection space) and / or absolute identification (e.g., confirming the same entity when it returns to the detection space, and / or direct identification of the entity by comparing with characteristics of known entities - for example determining that person “10105” has entered the detection space, and that person “10105” was last in the detection space four days ago; in another example determining that “Bob Smith” has entered the detection space based on known characteristics for “Bob Smith” that are available to the system). Example embodiments herein allow for the identification and / or labelling of entities, as well as utilizing that identification and / or labelling to determine and / or respond to events, and / or to provide notifications or alerts to users of the system. Example embodiments herein allow for the identification and / or labeling of actions by entities, as well as utilizing that identification and / or labeling to determine and / or respond to events, and / or to provide notifications or alerts to users of the system.Attorney Docket No. CURV-0003-WQ

[0036] Example embodiments of the present disclosure reference a qualia of vision related nouns, such as a vision space, of a visualization, or the like. As used herein, the qualia of a vision related noun references how the underlying result of the vision related noun is stored by the system. Where the vision related noun is stored as data directly imageable by humans, for example a picture data file, displayed picture data, a video data file, streaming video data, or the like, such data would be understood herein to be in a human type qualia of the vision information (or, in a human “qualia space”), even though, for example, a human could not actually see the picture data file. However, such data is in a ready format for display to humans, and accordingly could incur a risk that the data could be mismanaged, could include personally identifiable information, or the like. In another example, where the vision related noun is stored as data that is not imageable by humans, for example the determination of entity locations, entity model values, joint model values, entity posture values, and / or entity action values, such data would be understood to not be in a human qualia space, and / or may be referenced as in a wireless vision qualia space or similar terminology. Examples of the present disclosure use channel state information and / or other complex propagation data from configured messages to determine the wireless vision qualia space information, for example using a deep learning component against a base of channel state information, trained to detect entities, resulting in pattern recognition against that channel state information to determine the entity parameters set forth throughout the present disclosure. Examples of wireless vision qualia space data include determined entity related values that have been determined from the channel state information, and / or may include processed intermediates between the raw channel state information and the fully determined entity values, for example to allow some shifting of processing resource utilization, intermediate data storage to support wireless vision operations, and network communications, whether the shifting is a time shifting, and / or a shifting between computing devices. Accordingly, data that is sufficient to determine the entity values, and that has had an increment of processing from base channel state information to indicate that data of interest related to entities is actually in that sufficient data, represents wireless vision qualia space; where such data is sufficiently processed into data that can be readily displayed to a human, for example data in a standard format that would be likely to be understood as image data if released to a computing device or operating system that did not have context of the source of the data, then that data has shifted into a human qualia space.

[0037] Referencing Fig. 10, an example vision output of wireless vision systems herein is schematically depicted. At a top portion 1002, a visual view of the scene as it would be seen by a human is schematically depicted, with the scene built in the visual photonic domain. Three people standing in hallway can be observed, with one of the three people partially obscured by a wall.Attorney Docket No. CURV-0003-WQWhile this person can be seen visually, they are partially out of sight, and determination of posture or activity of the person would be difficult to determine. Additionally within about a step, the person will be out of view, therefore the person will disappear in a moment, or just now came into sight, depending on the direction of movement. In the bottom portion 1004, a wireless view of the scene is schematically depicted. The wireless view 1004 is converted into human visual qualia, for illustration. All of the people can be seen in the entirety, and a properly configured wireless vision system could determine postures and activities of the people, as well as the presence of metal objects of the like. Additionally, the wireless view 1004 is depicted in human visual qualia for clarity of the illustration, but the system could determine postures, activities, presence of metal object, and / or perform any other visual operations herein in wireless qualia space, allowing for the system to perform various monitoring and security functions without generating a human visual qualia representation, only doing so once an alert or event is indicated, and only according to scheduled responses for the alert or event. In the example of Fig. 11, a visual view 1002 of a scene is schematically depicted on the left side, and a wireless view 1004 of the scene is schematically depicted on the right side. The example of Fig 11 depicts a polygon based construction 1102 of the detected entity (a person, in the example), which may form a part of and / or be utilized instead of or in addition to other determinations such as an entity model determination, entity limb positioning determinations, etc. Referencing Fig. 12, a wireless view 1004 of a scene is schematically depicted with a number of detected entities in various postures as may be seen by a wireless vision system. The example entities maybe depicted based on determined postures, limb and / or joint models, and / or entity modeling parameters such as material characteristics, volume determinations, geometric determinations, distribution of mass, or the like. Limb positions (e.g., limb position 1202) are depicted as available from operations of the wireless vision system.

[0038] In certain embodiments, aspects of the present disclosure are referenced as a visualization. Except where the context explicitly indicates otherwise, a visual or visualization as used herein may reference a visualization in a wireless qualia space, a human qualia space, and / or both. Where a visualization is actually displayed to a user, that may include data that is stored in a human qualia space, and / or data in a wireless qualia space that is translated into a human qualia space for display purposes. Such data may be stored as human qualia space data after the display, and / or the human qualia space data may be deleted after the display operations are completed.

[0039] Referencing Fig. 1, an example procedure for providing entity vision operations utilizing a wireless vision system is schematically depicted. The example procedure includes an operation 102 to prepare and position a main vision device in a calibration location. The example operation 102 includes providing a calibration location that is similar to the target location (e.g., reference Fig. 2)Attorney Docket No. CURV-0003-WQfor which wireless vision is desired, where the main vision device includes a wireless transceiver having at least two fixed antenna banks and a wireless controller that provides channel state information (CSI) for communications between the fixed antenna banks, and / or for communications between the fixed antenna banks and antenna banks on auxiliary vision devices, between antenna banks on distinct auxiliary vision devices, and / or for communications between fixed antenna banks on auxiliary vision devices. In certain embodiments, the information for communications utilized to build the wireless vision (wireless vision data) of the present disclosure utilizes CSI, however any other type of information that includes sufficient information for the communications, including phase and amplitude data, as well as time stamp data, may be utilized instead of, or in combination with, available CSI. The devices should have antenna banks that are in a known position relative to each other, and the more precisely the locations are fixed, the more rapidly the wireless vision to entity vision learning will converge, and the greater precision that will be available in the entity vision operations performed in the real system after calibration. The calibration location should be covered by cameras with registered positions that can relate camera data to location data within the selected coordinate system, and where antenna banks of the various vision devices (e.g., main device and / or any auxiliary devices) are also registered in the selected coordinate system (and / or the coordinate systems are transformable between each other). Camera coverage for the calibration location should include coverage for any blind spot areas that will be of interest in the target location (e.g., behind furniture, low traffic areas, stairwells) and / or challenging areas (e.g., near large metal objects, behind a stainless steel appliance, etc.).

[0040] Embodiments herein describing the utilization of channel state information may additionally or alternatively utilize propagation data of any type for configured messages, regardless of the source or formal naming convention for the data. Accordingly, throughout the present disclosure, wherever channel state information is referenced, such references are understood to include propagation data of any type for configured messages.

[0041] In certain embodiments, channel state information as used herein may include channel frequency response (CFR) information. In certain embodiments, multipath interference causes multipath fading, which can have a frequency dependent component, and the CFR information characterizes multipath fading in the frequency domain, which allows the wireless vision system to enhance the vision with aspects that depend on vision aspects that have a frequency dependent characteristic (or spectral variability).

[0042] In certain embodiments, channel state information as used herein may include channel impulse response (OR) information. CIR characterizes multipath fading in the time domain, and is the time domain counterpart of the CFR information. Accordingly, CIR information may be utilizedAttorney Docket No. CURV-0003-WQin certain embodiments to enhance the vision with aspects that depend on vision aspects having a frequency dependent characteristic.

[0043] There are numerous frequency dependent aspects of photon response in the target location for vision, including without limitation: the material of entities, the composition of ambient air or other materials, the distance to objects, object reflection parameters, or the like. The described effects are not limiting, and some effects may be higher order effects (e.g., distance to the object enhances other frequency dependencies in the situation), or effects on frequency response of communications appear but the cause may not be well understood. The deep learning model and calibration operations as set forth herein, having access to frequency dependent data, including the CFR and / or the CIR, can nevertheless compensate for and / or utilize those effects in the vision system.

[0044] The example procedure includes an operation 104 to operate a deep learning algorithm to prepare entity vision operations. The example deep learning algorithm trains on entities identified by an expert user from the camera data (and / or utilizing conventional photo recognition operations on camera visual data to identify entities and label the training data), allowing the wireless vision controller to identify entities that are present in the wireless vision data. The example deep learning algorithm can readily determine the presence, category (e.g., person, furniture, structure, pet, etc.), posture (e.g., the configuration of entity joints and overall position of the entity), and / or activity (e.g., based on sequenced postures, and / or any other patterns in the wireless vision data that evidence particular activities based on labeled training activity) of entities in the target space. The target space includes the area surrounding any antenna pairs, generally in a spherical shell around the antenna banks and / or other antenna pairs. In certain embodiments, vision devices herein operate most efficiently on a single floor at a time (e.g., a floor of a building), and may be trained that way, such that the target space is a generally circular region around the antenna pairs, with a small blind region in very close proximity to the antenna banks (in certain embodiments, coverage in the blind region may be provided by auxiliary devices). The actual coverage range of a single device depends upon the material present in the target location, the entity vision operations of interest, the desired resolution for detecting entities and identifying entity posture and actions, or the like. In certain embodiments, coverage in a 10 foot radius or 30 foot radius can be readily achieved in most ordinary structures. In certain embodiments, coverage in a 100 foot radius can be achieved for low complexity structures or open spaces. In certain embodiments, coverage can readily be extended arbitrarily with auxiliary devices, and / or by utilizing increased transmission power (appropriately calibrated) and / or utilizing multiple transmission frequencies (also appropriately calibrated).Attorney Docket No. CURV-0003-WQ

[0045] Referencing Fig. 8, an example target location 902 is depicted with a wireless vision device 802 positioned above an entrance. The example wireless vision device 802 can readily determine ingress and egress operations of people or other entities (e.g., pets, shipments, etc.) to the building that includes the target location 902, as well as detect significant metal devices, as large metal devices can provide a distinct signal in the wireless qualia space. An embodiment such as that depicted in Fig. 8 provides a number of benefits over previously known systems, including providing a system that is inexpensive, can be operated without generating personally identifiable information, is not intrusive to persons in the area in the way that a camera would be, and can readily determine certain security risks such as large metal objects, containers, or weapons. Additionally, an embodiment such as that depicted in Fig. 8 cannot readily be obscured by simple measures and defeated in the way a camera could be, such as by a person putting on a mask, hiding a weapon in a backpack, and / or painting or otherwise obstructing a lens of a camera.

[0046] Embodiments herein are described using wireless communications, which may be in any frequency, including any 802.11 supported frequency (e.g., 5 GHz, 2.4 GHz, 60 GHz), ultrawide band (UWB), Bluetooth, software defined radio (SDR, which allows many different bands to be utilized), and / or mmwave and / or extreme high frequency (e.g., 24 GHz to 71 GHz, 30 GHz to 300 GHz, etc.). In certain embodiments, frequencies may be combined, for example utilizing mmwave within a given room, and standard WiFi frequencies throughout a structure, allowing for high resolution detection of entities in spaces of interest, as well as detection capability for extended regions such as stairwells, entry ways, or the like where high frequency signals may be attenuated. In certain embodiments, the deep learning algorithm further adjusts vision communication operations during the calibration operations, for example the communications between antenna banks, sequencing of communications, frequency progressions, power variations, or the like, to ensure that the desired wireless vision parameters can be met.

[0047] As utilized herein, wireless vision operations include operations to detect and differentiate wireless signals, including determining the source of the signal, the time, phase, and amplitude of the signal, and to differentiate where the signal was received. The entire set of wireless communications and inferences therefrom makes up the wireless vision, which creates a view of the world that is in a distinct qualia space from that intelligible by humans, but which nevertheless is translatable. As utilized herein, entity vision operations include operations to determine the presence, identity, posture, and / or any other aspect of an entity present at the target location, translated into a qualia space that is intelligible to humans. Thus, the deep learning algorithm utilizes the training data set to recognize the wireless vision aspects that translate to the entity vision aspects, and to allow aAttorney Docket No. CURV-0003-WQwireless vision controller to create an entity depiction for utilization in various operations throughout the present disclosure.

[0048] The example procedure includes an operation 106 to position a main vision device and / or auxiliary vision devices at the target location. In certain embodiments, the layout of devices at the target location should be consistent with the layout of devices at the calibration location. In certain embodiments, the calibration location may be the target location (e.g., calibrating at the location during operational downtimes of the target location), and / or further tuning calibrations may be performed at the target location even after the original calibration operations performed at the calibration location. The utilization of a calibration location is optional, and illustrates that embodiments can utilize a calibration location, but in certain embodiments, any or all of the calibration operations may be performed at the target location. The utilization of a calibration location and related operations can significantly reduce the amount of time for the wireless vision system to learn the space at the actual target location, and help ensure the system is properly configured (e.g., antenna arrangements, number and position of wireless vision devices, frequency selections of configured messages, message rates, etc.) to be successful at the achieving the intended purpose once at the target location, reducing what might otherwise be an iterative installation and testing of various parameters until the wireless vision goals are achieved.

[0049] The example procedure includes an operation 108 to gather wireless vision data, for example communications between antenna banks of the devices (e.g., the main vision device, and / or any auxiliary vision device if present), including performing selected communication operations and collecting the wireless vision data (e.g., CSI, phase, amplitude, timestamp, time-of-flight, and / or per-subcarrier values (e.g., signal-to-noise ratios) information). In certain embodiments, the procedure includes an operation 110 to perform entity vision operations, including translating the wireless vision into entity vision, locating entities in the target location, identifying entities in the target location, determining entity model coefficients (e.g., describing the entity material, size, and / or distribution of these), determining entity joint locations and orientations, determining entity postures, etc. Operation 110 includes performing any entity vision operations as set forth throughout the present disclosure.

[0050] In certain embodiments, the procedure includes an operation 112 to perform ongoing learning operations, for example by capturing the wireless vision data 108 and / or portions thereof, which may be utilized to keep the calibration of the vision system up to date (e.g., allows for improved entity vision operations as the location changes and / or as the conditions of the location change) and / or to improve vision operations over time. In certain embodiments, the operation 112Attorney Docket No. CURV-0003-WQincludes storing the data for a period of time, and periodically and / or episodically performing an update operation 114 to adjust the wireless vision processing of the associated wireless vision device.

[0051] Referencing Fig. 2, example dimensions of a calibration location that should be simulated relative to the target location are schematically depicted, as calibration location setup 202 aspects in the example of Fig. 2. The aspects that should be the same, and the required similarity between them, will depend upon the purpose of the wireless vision system and the selected entity vision operations. For example, where entity presence only is sufficient for the vision system, or for at least certain areas of the target location, then a looser match between the calibration location and the target location may be acceptable, for example a similar gross layout of obstacles and obstacle types, as well as similar distances to areas of interest. Where high resolution entity definition is desirable, such as matching model coefficients and / or joint positioning to specifically identifiable entities is desired, then the match between the calibration location and the target location should be a higher fidelity, and further significant camera coverage of all areas of interest is also desirable for calibration operations. Example and non-limiting match dimensions between the calibration location and the target location include, without limitation: obstacle types 204 (e.g., obstacle materials, shape, thickness, orientation, etc.); obstacle locations 206 (e.g., positioning between areas of interest, distance from antenna banks, etc.); entity types 208 (e.g., persons, furniture, specific objects, pets, etc.); entity locations 210 (e.g., locations within the space, including relative to obstacles); entity postures / activities 212 (e.g., entity postures of specific interest, and / or including postures that may be challenging such as laying down in an unexpected area, oblique positions, on a ladder or elevated position, and / or activities of interest which may further be labeled during training operations); metal object locations 214 (e.g., including locations of interest, such as metal objects in proximity to entities, and / or entities have certain postures or in certain locations); camera coverage 216 (e.g., ensuring camera coverage in the calibration location provides desired vision in the target location, such that the calibration operations build that vision into the wireless vision to entity vision conversion); and / or coordinate selections 218, reference locations, and / or transforms between these (e.g., ensuring selected coordinates from the calibration location transfer to the target location, ensuring positioning of the main wireless vision device and / or auxiliary devices matches in both locations, ensuring that significant reference locations are positioned the same (e.g., hidden areas, specific obstacles, specific areas of interest), and / or ensuring that coordinate systems utilized for various positions (e.g., entities, devices, and / or reference locations) are the same and / or can be transformed between each other). The example matching dimensions between the calibration location and the target location are non-limiting examples.Attorney Docket No. CURV-0003-WQ

[0052] Referencing Fig. 3, example and non-limiting entity vision operations 302 are schematically depicted. In certain embodiments, operations on Fig. 3 involve translating the wireless vision to entity vision, or translating from wireless vision qualia space to a human qualia space. In certain embodiments, one or more operations on Fig. 3 involve processing of information for entities that have been detected, where that processing may be performed directly from the wireless vision or from the entity vision (e.g., a response to an entity in a stairwell can be performed whether the entity is translated into and / or depicted in viewable entity space or not). In certain embodiments, one or more operations on Fig. 3 may include operations utilizing either wireless vision, entity vision, both, or a combination of these. An example operation includes an entity count 304 operation, allowing for the counting of a number of entities in a give location or space, and / or at a particular location of interest. In certain embodiments, the entity counting 304 operation may further include an identification of where entities are positioned, changes relative to a nominal value, expected value, or recent value, or the like. In certain embodiments, the number of entities may be quantitative (e.g., three persons in “Room 308’’), categorical (e.g., less than five, or “sparse” vs. “dense”, etc.), or a combination of these. An example operation includes an entity position 306 operation, for example a description of the position of an entity, facing or orientation, and / or position with respect to another entity (e.g., a person on a ladder, on a couch, leaning on a refrigerator, two rooms away from another entity, etc.) or structure (e.g., near stairs, in a doorway, in a stall, etc.). An example operation includes an entity categorization 308 operation, for example a description of the entity type (e.g., person, furniture, structure, pet, appliance, etc.) and / or basic information evident from the entity gross appearance (e.g., height, demographic information from entity mass and material constituency, etc.). An example operation includes an entity posture and / or activity 310 operation, for example a description of the entity posture or self-positioning (e.g., position of limbs and torso for a person, door open for a cabinet or refrigerator, and / or any other posture information that is of interest for the application), including, for example, determining whether the entity is at risk (e.g., unexpected lying posture, aggressive stance, defensive stance, unusual motion or lack of motion, etc.), and / or detecting activities of interest by the entity. An example operation includes an entity identification 312 operation, which can be an instance identification (e.g., tracking 314 and / or labeling (e.g., as an entity identifier 312 and / or entity categorization 308) a specific entity as they move through the target location, and / or identifying them again if they are briefly not detected within the target location) or an absolute identification (e.g., detecting a specific entity that returns to the target location after an extended period, which may still be pseudonymous, and / or making an absolute identification of the entity, for example based on a model-based fingerprint of the individual made during prior appearance in the target location, based on a database of matching information, and / orAttorney Docket No. CURV-0003-WQbased on information captured for the individual during calibration operations). The operations to match and / or track 314 individuals may be made using matching model coefficients (e.g., which will relate to the size, geometric shape, and / or material constituency of the individual), matching joint locations, and / or utilizing continuity determinations (e.g., tracking a moving user) or other rationality checks (e.g., logical limits on the speed of the entity in the space, gross comparisons of likely mass and / or geometric configuration of the entity, matching of movement habits or patterns, etc.). In certain embodiments, the matching and / or tracking 314 operations may include a confidence indicator, where certain operations performed may include a determination of the confidence of the matching and / or tracking operations in identifying the entity of interest.

[0053] An example operation includes an operation 316 for an alert, a notification, or other action related to the entity vision as an entity vision operation, which may include determining an aspect of interest related to an entity (e.g., based on a posture, activity, presence of a related metal entity, presence in an unexpected location, etc.), and providing a communication, for example to an external computing device such as a server 1308 and / or other external computing device 1310, making the information available to external users, monitors, administrators, security personnel, safety personnel, and / or systems related to these for further processing and / or action to be performed. Example and non-limiting entity vision operations include any operations set forth throughout the present disclosure, including at least providing a notification to a user (e.g., of a system utilizing the wireless vision device) based on the entity vision, alerting a user based on the entity vision, and / or performing an action based on the entity vision. The entity vision operation may be performed in response to the presence of an entity (e.g., a person, pet, detected object, etc.), the movement of an entity within the target area, an action of an entity in the target area, and / or any of these with regard to more than one entity, or the absence of an expected entity or entities.

[0054] Referencing Fig. 3, example and non-limiting entity vision operations 302 include one or more operations such as: determining and / or communicating an entity count 304 (e.g., within the target area, at a selected area, and / or of entities having a selected property); determining and / or communicating an entity position 306; determining and / or communicating an entity categorization 308 (e.g., type of entity, characteristic of entity, etc.); determining and / or communicating an entity posture and / or entity activity 310; and / or determining an entity identification 312 (e.g., including potentially tracking 314 an entity for continuity in the target area, recognizing the return of a specific entity to the target area, and / or identifying an entity in a pseudonymous and / or absolute manner).

[0055] Referencing Fig. 4, an example main vision device 400 (or wireless vision device) is schematically depicted. The example main vision device 400 includes two antenna banks 402, 404 in a fixed arrangement. The example antenna banks include three (3) orthogonally arrangedAttorney Docket No. CURV-0003-WQantennas on each bank, where communications in both directions between antennas of the banks form the wireless vision data, with phase, amplitude, time stamping, and / or CSI 408 utilized to form the data set, and a deep learning algorithm, and / or model derived therefrom, operating to convert the wireless vision data into wireless vision space (e.g., wireless vision qualia) to entity vision space (e.g., human vision qualia). The antenna banks 402, 404 may be mounted on a device providing a fixed relationship between the antenna banks 402, 404, allowing the main vision device 400 to be placed without performing any measurements or introducing additional measurement errors in determining the relative position between the antenna banks. The example main vision device 400 includes a wireless controller 406, which may be a standard controller such as utilized by a router, and which may be operated in a standard format (e.g., where the device inherently collects CSI and / or phase / amplitude information) and / or which may be modified to collect CSI and / or phase / amplitude information for signals passed between the antenna banks 402, 404. The system may include additional devices (e.g., reference Fig. 9, devices 912), with additional antenna banks, distributed throughout the target location depending upon the size of the target location, the position of obstacles throughout the space, and / or the desired resolution of wireless vision throughout the target location. Additional devices 912 are positioned in a known location relative to the main vision device 400 (or 904, in the example of Fig. 9), and / or positioned and then the position relationship is determined (e.g., by direct measurement, by camera determined positioning from cameras in known positions that can view the vision devices with sufficient resolution, skew correction, and / or other positioning determinations, etc., by position determination using dedicated communication signals, and / or by any other positioning mechanism). The example main vision device 400 includes a vision controller 412 that converts the wireless vision (e.g., represented in the phase / amplitude / CSI for wireless signals) into entity vision (e.g., utilizing the deep learning algorithm to convert wireless vision to entity vision), and provides the entity vision outputs 414 for further utilization in entity vision operations as set forth throughout the present disclosure. The vision controller 412 provides vision communications 410, which includes commands to send messages between antenna banks 402, 404 to create the wireless vision, which may include message frequency, sequencing, which antennas are sending and / or receiving, selected communication frequency (e.g., frequency of the photons utilized to perform the communications), and the like. The example vision communications 410 are depicted on the main vision device 400, but may additionally or alternatively be provided on and / or performed by auxiliary vision devices (e.g., 912). The vision controller 412 is depicted as a single device positioned on the main vision device 400 for clarity of the present description, but the vision controller 412 may be a single device or a distributed device, and may be positioned, in whole or part, on at least one or more of: a separate controller on the main vision device (e.g., as depictedAttorney Docket No. CURV-0003-WQin Fig. 4), co-located with the wireless controller 406, on a separate controller of an auxiliary vision device 912, co-located with a wireless controller of an auxiliary vision device 912, and / or on a separate device (e.g., a local computer at least selectively in communication with the main vision device such as 1310, and / or on a cloud device and / or server such as 1308).

[0056]

[0057] Referencing Fig. 9, an example and non-limiting arrangement of a system positioned at a target location 902 is schematically depicted. The example depicts a main vision device 904 somewhat centrally located in the target location 902 (this arrangement may be convenient but is not required), and two auxiliary devices 912 to extend the vision range and / or provide enhanced vision in areas of interest. In the example of Fig. 9, three areas of interest are depicted, which may be any type of area, such as an entry area 910 (e.g., Area 3), a remote area 908 where camera coverage may be undesirable, unavailable, and / or too expensive to provide (e.g., Area 2), and a private area 906 where cameras may be undesirable and / or unavailable, but entity vision (e.g., the presence, number, and / or postures of entities, and / or certain activities - e.g., where certain activities of interest can be recognized but other activities would remain undetected or undefined) may be desired (e.g., a bathroom, living space, locker room, etc., such as Areal in the example). The vision systems in the examples of the present disclosure allow for privacy protected vision into spaces without requiring the infrastructure to position cameras and / or with a reduced overhead to determine and manage privacy considerations, and allowing operations to be performed in wireless vision qualia space except as needed, minimizing risks, costs, and intrusion of the system.

[0058] Example systems may perform any one or more of the following operations, without limitation:

[0059] Building a coefficient based model of entities, reflecting the volume, geometry, and / or materials of entities within the wireless vision space;

[0060] Building a joint model of entities, and / or including joint positions, of entities within the wireless vision space;

[0061] Determining a posture of entities within the wireless vision space, including for example orientation of the entity, positioning of joints and / or limbs, and / or gross positioning (e.g., standing, leaning, laying down, etc.);

[0062] Determining a relationship between entities within the wireless vision space, for example determining a person entity is sitting on a chair entity, etc.;

[0063] Detecting significant metal objects and / or entities in the space;

[0064] Determining a relationship (e.g., proximity, presence in a bag, etc.) of significant metal objects with other entities in the space;Attorney Docket No. CURV-0003-WQ

[0065] Determining where people are located within the space, where people are moving to within the space (and / or where they are exiting the space), dwell times within the space, and / or tracking presence and movement within the space during a given trip through the space and / or during two or more trips through the space;

[0066] Determining that entities are present at a sensitive, hazardous, unusual, impermissible, and / or unexpected location within the space;

[0067] Providing a 360 degree field of view around a main vision device within the space (and / or a 90 degree (e.g., internal corner position), 180 degree (e.g., hall or wall position), and / or 270 degree (e.g., external comer position) field of view);

[0068] Utilizing an expert system, reinforced learning operation, and / or operating a large language model (LLM) for entities in the wireless vision to be labeled, tagged, and / or displayed in the entity vision;

[0069] Utilizing an expert system, reinforced learning operation, and / or operating an LLM for actions of entities in the wireless vision to be labeled, tagged, and / or displayed in the entity vision;

[0070] Tracking entities in the space to ensure that they are in a safe place, and / or to ensure that all of the people leave in a hazardous condition (e.g., during a fire or fire drill);

[0071] Tracking entities in the space for unusual or suspicious activity (e.g., individuals gathered in an unusual location, performing a hazardous activity such as walking into a blind comer where another entity such as a forklift may be encountered);

[0072] Tracking entities in the space that may need assistance (e.g., an individual lying in a bed at unexpected times, an individual that has fallen in the shower or bathroom, an individual that is unusually still or otherwise in a posture or activity indicating stress, etc.); and / or

[0073] Tracking individuals performing suspicious actions and / or having significant metal objects that may indicate a hazard, for example at entryways and / or otherwise throughout the target location.

[0074] In certain embodiments, an antenna subsystem for a wireless vision device comprises two or more fixed antenna banks mechanically registered to one another and to the device chassis to establish a stable and known baseline for phase, amplitude, and timing measurements used to generate wireless vision data. Each antenna bank may include one or more patch antenna elements 512 and be arranged in a defined geometry to provide controlled spatial coverage and controlled inter-element isolation. In certain embodiments, patch antennas 512 combined with a waveform cone may be referenced as a conical antenna, and / or other antenna types may be utilized such as a conical or spiral antenna element. In certain embodiments, any type of antenna and / or arrangement may be utilized that is capable to propagate EM radiation of selected frequencies into the target locationAttorney Docket No. CURV-0003-WQ(and / or specified area) and receive returns which can be mapped between antennae to determine propagation effects of the environment on those photons.

[0075] Fig. 5 shows a schematic view of one example of a representative embodiment used for illustration. Fig. 5 shows a two-bank configuration employed in which each bank includes four patch antennas 512, thereby creating a 4x4 bidirectional link matrix between the banks that yields sixteen independent transmit-receive paths. More generally, any number of banks and any number of antennas per bank may be utilized, with the resulting link matrix comprising the Cartesian product of elements across distinct banks and optionally within banks.

[0076] Fig. 5 shows an example mechanical layout of an antenna array system designed for wireless vision operations, which may be referenced herein as a wireless vision device or similar terminology. The example embodiment of Fig. 5 may embody all or a portion of a wireless vision system, and elements of the wireless vision system may be distributed, in whole or in part, on other devices (not shown) of a system, for example with elements of a radio transceiver, antennae and / or banks of antennae, controllers, or the like, positioned within a same housing as the wireless vision device, positioned in a housing coupled to a housing of the wireless vision device, and / or positioned separately from the wireless vision device and at least selectively operatively and / or communicatively coupled to the wireless vision device, and forming a wireless vision system. The layout incorporates several components that contribute to the functionality and structural integrity of the system, housing access 502 provides an access point to decouple an antenna housing 602 (e.g., reference Fig. 6) from a control housing 606, and / or to serve as a mounting point for the wireless vision device. Referencing Fig. 7, an example bottom view of a wireless vision device depicts the antenna housing 602 and an access cover 608 that can be removed to allow access into the housing, and to the housing access 502.

[0077] The mounting base 506 provides a platform for the antenna array system, and / or forms a part of the antenna housing 602. In certain embodiments, the mounting base 506 may be constructed from materials designed to offer durability and resistance to environmental factors, such as temperature variations and mechanical stress. In one implementation, the mounting base 506 may include, and / or be closely coupled to, a printed circuit board (PCB) that provides both mechanical support for the antenna cones 504, and provides electrical connectivity for the antennae, and power and communication coupling for controllers of the wireless vision system. When implemented as a PCB, the base may include controlled-impedance traces, ground reference planes, and / or shielding features to maintain element-to-element isolation while contributing to structurally rigidity of the substrate for the array assembly.Attorney Docket No. CURV-0003-WQ

[0078] Support rods 508 may be integrated into the design to provide additional structural support and rigidity to the antenna array, and to couple cover plates of the antenna housing 602. These rods 508 help maintain the alignment of the components and prevent deformation or displacement during operation or installation. In some implementations the support rods 508 may provide and offset and / or support structures for a cover. The outer frame 510 encloses the antenna array system, offering protection and structural integrity, and / or forming a part of the antenna housing 602. The outer frame 510 enables the incorporation of the antenna array system into different deployment environments, such as ceilings or walls by providing a suitable mechanical structure for fastners, brackets, and the like.

[0079] One or more of the antenna elements may include a power circuit feed structure that imposes a defined polarization state, for example in a hardware arrangement where the orientation of powering and detection circuits coupled to each antenna (e.g., as formed on a PCB, or other arrangements) define and / or contribute to a polarization of the messages sent and / or received by the given antenna. In certain embodiments, the polarization of antennae during operations may be adjusted in real time, for example through the utilization of multiple circuits that may be enabled or disabled, and / or through dynamic control of circuits coupled to the antennae. In one implementation, a T-shaped feed or an equivalent asymmetric feed network excites orthogonal modes of the patch in a prescribed phase relationship to produce a linear or rotated linear polarization at the element aperture. Across the bank, the polarization axes of adjacent elements may be rotated relative to one another by selected angular offsets so that the polarization vectors are separated in three-dimensional polarization space. In certain embodiments, maximum polarization differences may be imposed, for example in a 4 antennae banks such as that depicted in the example of Fig. 5, polarization separation of 90 degrees may be applied to provide maximum de-coupling of signals and ease of keeping communications between antennae separate. Rotational offsets may be evenly distributed around a 180-degree span for linear polarization states or distributed over a 360-degree span for implementations that produce elliptical or circular components. In certain embodiments, sufficient polarization separation is provided that is not maximal - for example based on simulation and experience, it is believed that a polarization separation of about 10 degrees between antennae provides for sufficient separation that signals between antennae can be readily separated. In certain embodiments, separation between antennae is alternatively and / or additionally imposed by other techniques, for example utilizing time based separation (which may limit signal rates, and accordingly system responsiveness, resolution of entities identified in the wireless vision qualia, and / or reducing the number of features such as model coefficients and / or entity joints that can be resolved, and / or reducing the number of distinct frequencies that can be operated in a given timeAttorney Docket No. CURV-0003-WQperiod to support wireless vision operations), and / or utilizing intelligent signaling separation (e.g., applying signatures, metadata, and / or other signal based aspects to messages to allow for separation of signal sources and destinations between antennae during processing operations of the received signals). The deliberate polarization diversity among elements may be used to reduce mutual coupling, promote channel decorrelation, and increas the separability of signals received from and transmitted to different elements when the wireless controller cycles through antenna pairs during channel sounding.

[0080] In some embodiments, each antenna may include a cone structure mechanically registered to the PCB and aligned to the normal of the radiating surface. The cone may be configured to act as a controlled-aperture waveguide and / or a directional shield that narrows the element beam to a defined field of view and attenuates sidelobe energy outside of a designated solid angle. In certain embodiments, the circular antenna cones 504 may be fabricated from materials selected to provide desired electromagnetic, mechanical, and environmental performance while meeting manufacturing and cost constraints. Suitable materials include, without limitation: metallized polymers, conductive sheet metal, plated composites, hybrid cones with absorptive liners (e.g., where a conductive shell is combined with RF absorber materials), dielectric cones with frequency-selective surfaces (FSS), materials with surface treatments (e.g., conductive paints, electroless or electroplated copper / nickel, or sputtered aluminum), and / or the like. Each cone 504 may be mechanically registered to the corresponding board supported elements of the antenna to maintain alignment and stability during operation.

[0081] In the example of Fig. 5, two antenna banks are positioned at a fixed baseline distance and orientation relative to one another so that phase and amplitude measurements taken along different inter-bank paths are repeatable from deployment to deployment. The baseline is selected as a function of operating frequency, desired angular resolution, and installation constraints. The example of Fig. 5 includes two banks of four antennae, in a non-limiting arrangement. Antennae driver controllers 1306 (e.g., reference Fig. 13) in a single system on a chip (SOC) are commercially available that can drive 4 antennae, making a bank of four antennae a convenient and low cost arrangement. However, driver controllers 1306 can be created for any number of antennae, and / or in certain embodiments more than one driver controller 1306 may be utilized to support an antennae bank.

[0082] Referring again to Fig. 5, the antenna subsystem may be integrated with a radio board that houses two radio chipsets capable of operating simultaneously across the device’s selected bands. In one embodiment, Wi-Fi chipsets and their associated front-end modules operate with up to 160 MHz instantaneous bandwidth, enabling a dense set of subcarriers for fine-delay and frequency-selectiveAttorney Docket No. CURV-0003-WQchannel measurements. The PCB stack-up provides controlled impedance feed lines to each patch element, incorporates ground reference planes to maintain element-to-element isolation, and includes via fencing and shielding features around the feed networks and active components to limit parasitic coupling into the radiators. The radio board and antenna PCB(s) are mechanically coupled through precision standoffs and alignment features that preserve the rotational orientation of each patch radiator and the coaxial alignment of each cone structure.

[0083] The antenna element spacing within each bank may be selected to reduce mutual coupling while reducing array aperture size. Spacing may be established as a function of the patch’s effective aperture and the cone’s flare and height. The cones enable closer element placement by confining each element’s near-field energy and by shielding adjacent feeds from direct coupling paths.

[0084] In operation, the controller sequences transmissions and receptions across the 4x4 inter-bank matrix so that each patch in one bank communicates with each patch in the other bank in both directions. In embodiments, temporal scheduling may be used to separate closely spaced or copolarized paths where additional isolation is desired, and frequency hopping across available bands may be employed to further diversify the set of measured channels. The combination of polarization diversity, cone-based pattern control, and fixed inter-bank geometry stabilizes the channel measurements while mitigating self-interference.

[0085] While certain embodiments employ four elements per bank due to chipset constraints, the structural and electromagnetic principles extend to larger banks. As the element count increases, polarization angle assignments and cone geometries may be adjusted to preserve isolation and coverage properties. The patch elements may be tuned for a variety of bands, including 2.4 GHz, 5 GHz, and 6 GHz Wi-Fi bands, as well as other frequencies where regulatory and hardware support is available. Multi-band implementations may employ stacked patches, dual-resonant patches, or multiple radiators per element position, with cone dimensions proportioned to accommodate the highest operating frequency while maintaining acceptable performance at lower bands through tailored interior profiles or frequency-selective surfaces.

[0086] During manufacturing and calibration, each device may undergo a factory alignment procedure in which return loss, isolation, and composite pattern characteristics are measured and compared to specified thresholds. In some cases, per-element calibration coefficients may be stored to compensate for small phase and amplitude deviations introduced by manufacturing tolerances or by minor installation variances. The coefficients may be subsequently used by the wireless controller to normalize channel measurements.

[0087] Fig. 6 shows a side view and a front view of the antenna enclosure 602, a secondary enclosure 606 (e.g., a controller enclosure) where various controllers and / or circuits are housed toAttorney Docket No. CURV-0003-WQsupport operations of the wireless vision device, and heat fins 604 that support heat rejection from the wireless vision device. In certain embodiments, the controller enclosure 606 may include mounting features, such as bolt holes and / or a mounting bracket, to provide for convenient placement of the wireless sensing device at a selected location and orientation. The antenna enclosure 602 and secondary enclosure 606 provide housing for the internal components of the assembly, providing protection against environmental factors. The enclosures 602, 606 are configured to accommodate hardware, such as antenna banks, controllers, and other electronic components described with respect to Fig. 5. The example antenna enclosure 602 is depicted with a smooth, planar surface, which may include openings or attachment points for securing the device to the secondary enclosure 606. The secondary enclosure 606 is configured to facilitate the secure installation of the device enclosure 602 onto a surface, such as a wall or ceiling.

[0088] In certain embodiments, the device illustrated in Fig. 5 is adaptable to a variety of mechanical and electromagnetic configurations to accommodate deployment environments and sensing objectives. The antenna banks, element count, cone geometries, and enclosure form factor may be selected and scaled to tailor coverage, isolation, and channel richness without departing from the principles disclosed herein. By way of example and not limitation, the device may be configured as a ceiling-mounted unit in which the antenna banks are oriented to direct cone-shaped beams downward, producing a controlled interaction footprint beneath an entry way to monitor ingress and egress, count individuals, and detect objects. In another example, the device may be adapted for wall mounting along a corridor, with cone geometries adjusted to form an elongated lobe aligned with the hallway axis to track motion along a path while suppressing reflections from adjacent rooms. In yet another example, the device may be arranged in a room-centric configuration, with lobes distributed azimuthally to provide near-omnidirectional coverage for open areas such as lobbies or cafeterias.

[0089] Fig. 8 shows an example configuration wherein target location 902 (and / or specified area) is a building entrance / exit equipped with a ceiling-mounted device 802, which is part of the wireless vision system. The device 802 is positioned above the entrance / exit to monitor the area and perform entity vision operations as described in the present disclosure. The placement of the device 802 provides coverage of the entrance / exit area, enabling the detection and tracking of entities as they approach, pass through, or exit the building.

[0090] The ceiling -mounted device 802 incorporates antenna banks and a wireless controller, which are configured to collect wireless vision data, such as channel state information (CSI), phase, amplitude, and timestamp infomation. The device 802 utilizes this data to perform entity vision operations, including detecting the presence, position, and movement of entities within the monitored space. The device 802 is capable of identifying entities, determining their posture, and detectingAttorney Docket No. CURV-0003-WQactions or events of interest, such as unusual activity or the presence of significant metal objects. The device 802 is designed to provide a controlled field of view, focusing wireless signal coverage on the area beneath the exit.

[0091] The ceiling -mounted device 802 is enclosed in a protective housing 602, 606, which shields the internal components of the device from environmental factors and provides structural integrity. The housing 602, 606 is designed to blend seamlessly with the building's architecture, ensuring a discreet and unobtrusive installation.

[0092] In a top-down entryway embodiment, the cones collimate energy into a downward-pointing lobe so that the half-power coverage footprint on the floor forms a controlled region of approximately three meters by three meters directly below the device, with the cone angle and height selected to achieve the desired footprint and to suppress energy in directions that would cause device-to-device overlap or ceiling reflections.

[0093] In certain embodiments, the device operates by transmitting probe signals among antenna elements arranged in banks and measuring the resulting wireless channel responses to form a wireless vision of a monitored space. A controller sequences inter-bank transmissions using wideband pilots and tone sequences across selected bands and time slots. Receiving elements in opposite banks capture complex channel state information, including per-subcarrier amplitude and phase, from which impulse responses, delay spreads, Doppler signatures, and polarization-dependent attenuation are derived. Using the fixed geometry and calibration of the banks as spatial references, the controller transforms these measurements into entity-level outputs (e.g., presence, position, posture, and / or object signatures).

[0094] In operation, a controller may be configured to sequence transmissions and receptions among elements across distinct banks according to a pairing schedule. For a first bank having NAelements and a second bank having NBelements, the controller may be configured to form NAxA^fiinter-bank links in a forward direction and NBX NAinter-bank links in a reverse direction, producing a Cartesian product of transmit-receive paths that sample the monitored volume from multiple angles and polarizations. The transmitted probe signals include wideband pilots spanning an instantaneous bandwidth (for example, orthogonal frequency division multiplexing pilots up to 160 MHz), tone sequences that exercise discrete subcarriers, and, in certain embodiments, frequencyhopped bursts across available bands to diversify delay and reflection profiles. Temporal scheduling and power control may be adjusted to isolate closely spaced or co-polarized paths.

[0095] Receiving elements in an opposite bank measure quantities sufficient to reconstruct a wireless channel response, including complex channel state information comprising per-subcarrier amplitude and phase, packet timestamps, and scheduling metadata that identify the transmitting bankAttorney Docket No. CURV-0003-WQand element and the receiving bank and element for each sample. From these measurements, impulse responses, delay spreads, and multipath structures are derived. The fixed geometry of the banks provides known baselines and orientations for each inter-bank path.

[0096] The antenna banks define a structured set of propagation paths used to illuminate and sample the monitored volume while providing stable geometric and polarization references. The foregoing description is not limited to a two-bank configuration. Any number of banks and any number of antennas per bank may be utilized. In certain embodiments, three or more banks are arranged to increase angular sampling and volumetric coverage, and the controller generalizes pairing to all distinct bank pairs and, where desired, selected intra-bank paths. In a representative embodiment used for illustration, a two-bank configuration is employed in which each bank includes four patch antennas, thereby creating a 4x4 bidirectional link matrix that yields sixteen independent transmit-receive paths. More generally, the pairing schedule, polarization assignments, cone geometries, and bank separations are selected to preserve isolation and coverage properties as element counts and bank counts are scaled.

[0097] In certain embodiments, artificial intelligence is employed to transform wireless communication measurements into entity-level outputs by learning mappings from propagation features to human-interpretable descriptors. An Al subsystem may be integrated with or communicatively coupled to the controller and is operable at multiple stages of a processing pipeline.

[0098] In one aspect, the system constructs a training dataset by collecting synchronized wireless channel measurements and ground-truth annotations obtained during calibration, simulation, or co-registered sensing. An annotation mechanism associates positions, postures, activities, and object labels with time-aligned channel samples. In some embodiments, optical data, if present, is utilized only during calibration to generate labels and is not utilized during operational inference. The dataset is partitioned into training, validation, and test subsets, and data augmentation is applied to increase diversity, including perturbations to power, interference, and multipath conditions representative of deployment environments.

[0099] In another aspect, the Al subsystem comprises one or more neural network models configured to ingest features derived from per-link measurements, such features including, without limitation, per-subcarrier amplitude and phase, delay profiles, Doppler spectra, polarization contrasts, and stability metrics. Architectures may include at least one of: convolutional neural networks operating on frequency-time or Doppler-time representations; recurrent or sequence models configured to capture temporal dependencies across frames; graph neural networks that represent links as nodes or edges parameterized by bank geometry; and multitask networks thatAttorney Docket No. CURV-0003-WQjointly predict occupancy, position, posture, and object signatures with shared encoders and task-specific decoders.[000100] In operation, the Al subsystem maps input features to spatial outputs expressed in a defined coordinate frame established during installation. In one embodiment, a model produces an occupancy grid with per-cell presence probabilities. In another embodiment, a model regresses continuous coordinates and uncertainties for one or more entities. Loss functions for training include cross-entropy losses for occupancy estimation, regression losses for coordinate prediction, and auxiliary consistency losses that penalize unphysical motion and enforce reciprocity between forward and reverse links.[000101] In certain embodiments, the Al subsystem further comprises classification heads operable to output posture classes, activity labels, and object indicators associated with significant metal content or distinctive shape and density signatures. The classification heads are trained with class-balanced or focal losses to address class imbalance. For object detection, the subsystem exploits polarization-dependent attenuation and micro-Doppler cues to distinguish earned items from background clutter.[000102] Temporal association and tracking are performed by an Al-assisted data association mechanism that links per-frame detections into trajectories. A learned affinity model generates association scores based on spatial proximity, motion continuity, and feature similarity, and a recurrent tracker refines trajectories and suppresses spurious detections. Instance embeddings may be produced to maintain continuity for entities during brief occlusions or low-signal intervals.Confidence estimates are generated for each track and utilized to gate alerts and downstream actions.[000103] To maintain performance across sites and over time, the subsystem may include domain adaptation and continual learning. Techniques include transfer learning with limited labeled data from a new environment, adversarial domain alignment to reduce feature drift, and self-training with high-confidence pseudo-labels. Ongoing learning may be executed during off-peak intervals using stored datasets, subject to safeguards that constrain updates to preserve validated behavior and regulatory compliance.[000104] Al models may be deployed on-device, on an edge server, or in a cloud environment. In certain embodiments, compressed or quantized models execute locally to achieve low latency, with optional escalation to higher-capacity models running remotely for detailed analyses. Cascaded or ensemble configurations may be employed, wherein a lightweight detector triggers a more complex model upon detection of specified conditions.[000105] In certain embodiments, a wireless vision system is embodied at least in part as a wireless communication device comprising a radio transceiver and a plurality of antennas arranged in at leastAttorney Docket No. CURV-0003-WQtwo spatially separated banks. The banks are mechanically registered to one another and to a device chassis so that the separation distance, relative orientation, and polarization references are fixed and repeatable across deployments. Each bank includes one or more radiating elements, such as patch antennas, dipoles, slots, or multiband radiators, optionally integrated with pattern-shaping structures that collimate energy into selected lobes. The radio transceiver is configured to transmit and receive non-visual frequency photons across one or more designated bands, which may include, by way of example, 2.4 GHz, 5 GHz, and 6 GHz Wi-Fi bands, ultrawideband channels, 60 GHz or other millimeter-wave bands, Bluetooth channels, software-defined radio bands, and combinations thereof. Front-end modules provide filtering, switching, and gain control appropriate to the selected bands and support bidirectional operation with sufficient instantaneous bandwidth to resolve multipath delay and frequency selectivity within the monitored environment.[000106] A controller is in communication with the radio transceiver and with the plurality of antennas and is configured to orchestrate channel sounding and measurement collection. The controller selects antenna pairs from among the plurality of antennas, preferably across different banks to exploit the fixed baselines and to increase angular diversity, although in certain embodiments intra-bank paths are also exercised. Pair selection may follow a deterministic schedule, a pseudorandom sequence, or an adaptive policy that reacts to interference conditions, desired spatial sampling density, or application-specific objectives. For each selected pair, the controller instructs the radio transceiver to transmit configured messages between the antennas over one or more non-visual frequency channels. The configured messages include probe waveforms suitable for channel estimation, such as orthogonal frequency division multiplexing pilot frames spanning an instantaneous bandwidth up to 160 MHz or greater, tone sequences applied to discrete subcarriers, and frequency-hopped bursts scheduled across available bands to diversify delay and reflection profiles. Temporal slotting, duty cycling, and transmit power levels are controlled to maintain regulatory compliance and to reduce self-interference when closely spaced or co-polarized paths are exercised in succession. The utilization of configured message herein references signals passed intentionally between antenna pairs having selected characteristics. The configured messages are not typically utilized to convey information as in a typical network message, although typical network messages could be utilized for at least a portion of some of the configured messages, and / or typical network messages may be monitored and utilized as a part of the vision system, with dedicated configured messages for vision operations utilized to fill in gaps in the interrogation plan for the targeted location. In certain embodiments, all of the configured messages are dedicated vision messages, and are not utilized to perform or support normal network communications.Attorney Docket No. CURV-0003-WQ[000107] For the configured messages, the system obtains channel state information 1902 (e.g., reference Fig. 19) that comprises complex propagation descriptors sufficient to characterize the propagation of the non-visual frequency photons through a specified area. These descriptors include, without limitation, per-subcarrier values 1904 that encode amplitude 1908 and phase 1906 across the occupied bandwidth, and that can be utilized to determine signal-to-noise ratios that provide vision information to the wireless vision system; packet timestamps 1912 and element identifiers that map each sample to the transmitting and receiving antennas; phase-coherent time-of-flight 1910 or time-of- arrival estimates derived from wideband pilots; and derived quantities such as impulse responses, delay spreads, Doppler spectra, polarization-dependent attenuation, and path stability metrics. Reciprocity may be exploited by measuring both forward and reverse links between a given antenna pair to validate phase coherence and to compensate for transceiver asymmetries. Calibration coefficients obtained during manufacturing and installation are applied to normalize residual amplitude and phase offsets among elements so that inter-bank comparisons remain consistent.[000108] The controller processes the channel state information to infer entities within the specified area, including at least the presence and location of one or more entities. Processing may include digital signal processing to denoise measurements, to align phases across banks, and to compute feature vectors from the complex descriptors. Features may include subcarrier-wise fading profiles, temporal variations in phase and amplitude indicative of micro- and macro-motion, Doppler signatures, path perturbations caused by objects with significant metal content, and polarization contrasts associated with material composition and orientation. In one embodiment, the controller executes a trained machine learning model, such as a deep neural network, that maps the feature vectors to entity-level outputs comprising presence probabilities, estimated positions in a defined coordinate frame, and confidence measures. In other embodiments, the processing includes model-based estimators that fuse time-of-flight, baseline geometry, and multipath structure to localize entities, optionally constrained by environmental priors learned during calibration. The processing may further classify entities by category and, where applicable, identify objects carried by or proximate to entities when the signatures of those objects are separable from the background.[000109] The system generates a view that is a visualization of the entities in the specified area based on the channel state information and the inferred outputs. The view may be rendered locally on the device, on a connected client, or on a remote server, and may comprise a two-dimensional or three-dimensional depiction of the monitored space with markers indicating entity locations, trajectories, or postures, along with confidence levels and alert designations. Coordinates in the visualization correspond to the spatial reference established during installation, which may be registered to a building map or to a local reference frame. In certain embodiments, the visualizationAttorney Docket No. CURV-0003-WQis updated at a rate sufficient to track motion in real time, and event logic triggers notifications when entities enter or exit zones of interest, when counts exceed thresholds, or when signatures consistent with significant metal objects are detected.[000110] The foregoing operations are supported by the mechanical and electromagnetic configuration of the antenna banks. By pairing elements across banks, the controller exploits known baselines to sample the monitored volume from multiple angles and polarizations, enhancing sensitivity to perturbations introduced by entities and increasing the separability of distinct paths. Pattern-shaping structures, such as cones coupled to patch radiators, confine the interaction volume and attenuate sidelobes, thereby stabilizing measurements and improving coexistence when multiple devices are installed. While a two-bank arrangement provides a compact embodiment, the principles extend to three or more banks to increase angular sampling and volumetric coverage. Likewise, any number of antennas per bank may be employed, with the pairing schedule and processing pipelines scaled accordingly.[000111] In a representative deployment, the device is mounted on a ceiling above a building entrance, with cones oriented to direct energy downward and to produce a controlled footprint on the floor beneath the device. The controller cycles through inter-bank antenna pairs, acquires channel state information across the configured bands, and infers the presence and location of people passing beneath the device. The generated view presents ingress and egress counts and trajectories in the entry region and may further annotate detections associated with significant metal objects. In other deployments, the device is adapted for wall mounting along a corridor with elongated lobes aligned to the hallway axis, or for room-centric coverage with lobes distributed azimuthally to provide near-omnidirectional sensing. Across these configurations, the radio transceiver, antenna banks, and controller cooperate to transmit and receive non-visual frequency photons, to measure and interpret the resulting channels, and to convert wireless vision into a human-intelligible visualization of entities within the specified area.[000112] Referencing Fig. 13, an example wireless vision system consistent with embodiments of the present disclosure is schematically depicted. A Wi-Fi vision device 1300 houses a controller 1302, two antenna banks 1304, and antenna controllers 1306 operatively coupled to the respective banks. In operation, the controller 1302 orchestrates configured message transmissions among selected antenna elements of the banks 1304, while the antenna controllers 1306 implement per-element driving, switching, polarization control, and timing to support channel sounding and measurement collection. Within the device 1300, the controller 1302 exchanges control and measurement data bi-directionally with each antenna controller 1306 and maintains coordinationAttorney Docket No. CURV-0003-WQacross the antenna banks 1304 to acquire propagation information sufficient for wireless vision processing.[000113] The system further includes an external controller 1308 in communication with the device 1300; the external controller 1308 may be a cloud server, a web portal, and / or any other connected computing device operatively linked to the remainder of the system. In certain embodiments, the external controller 1308 hosts one or more operations otherwise performed by the controller 1302, including without limitation wireless-vision feature extraction, entity inference, model training or updating, fleet configuration, data storage, or alert orchestration.[000114] A display controller 1310 is also in communication with the device 1300 and, in some embodiments, with the external controller 1308, and may be embodied as a computing device of any type, including a mobile device, laptop, web portal access, and / or a computing device connected to the rest of the system, and may be associated with any stakeholder of the system (e.g., operators, administrators, safety personnel, or auditors). Without limitation, any visuals or displays that are, fully or partially, based on human visual qualia may be provided to the display controller 1310, including rendered views of a specified area, entity markers, trajectories, posture annotations, alert banners, and confidence indicators generated from wireless vision outputs.[000115] Again referencing Fig. 13, a Wi-Fi vision device 1300 includes a radio transceiver in communication with a plurality of antennas arranged in at least two spatially separated banks 1304 and a controller 1302 operable to orchestrate wireless- vision operations. In operation, the controller 1302 selects antenna pairs from among the plurality of antennas and transmits, via the radio transceiver, configured messages between the selected antenna pairs over one or more non-visual frequency channels. For these configured messages, the system obtains channel state information that characterizes propagation of non-visual frequency photons through a specified area, including, in some embodiments, parameters such as per-subcarrier values, phase, amplitude, time-of-flight, and timestamps. The controller 1302 processes the channel state information to infer entities within the specified area, including at least presence and location of one or more entities, and generates a view that is a visualization of the entities in the specified area based on the channel state information, produced without reliance on visible-light image data.[000116] In certain embodiments, the antenna banks 1304 comprise two banks of three linear antennas in each bank. In other embodiments, the antenna banks 1304 comprise two banks of at least two omnidirectional conical antennas in each bank; in a representative configuration, the banks comprise two banks of four omnidirectional conical antennas, supporting selection from sixteen cross-bank antenna pairs. A driving circuit (e.g., implemented within antenna controllers 1306) may apply a selected polarization to each omnidirectional conical antenna to increase decoupling amongAttorney Docket No. CURV-0003-WQradiation patterns, and antennas of a given bank may be separated by a polarization margin, which may include a polarization difference of at least 10 degrees. Corresponding antennas of each spatially separated bank may be operated in a matching polarization region to facilitate cross-bank pairing.[000117] The controller 1302 may transmit configured messages at a plurality of photon frequency ranges, which may span at least the radio and microwave spectrum. In various embodiments, the plurality of photon frequency ranges comprise frequency values between 1 MHz and 1 GHz (inclusive), between 104Hz and 10loHz, or between 105Hz and 109Hz. Each of the at least two spatially separated banks may be driven by independent driving circuits, and the driving circuits may comprise solid-state circuits. In compact deployments, the at least two spatially separated banks are present within a same housing of the wireless vision device 1300, with the radio transceiver and controller 1302 co-located to support low-latency acquisition and processing of channel state information.[000118] Again referencing Fig. 13, a Wi-Fi vision device 1300 includes a radio transceiver in communication with antenna banks 1304 and a controller 1302 that cooperates with antenna controllers 1306 to orchestrate wireless-vision operations for creating a Wi-Fi vision based entity model. In operation, the controller 1302 selects antenna pairs from among the antennas of the banks 1304 and transmits, via the radio transceiver, configured messages between the selected antenna pairs over one or more non-visual frequency channels. For these configured messages, the system obtains channel state information that characterizes propagation of non-visual frequency photons through a specified area and processes the channel state information to create, for one or more entities within the specified area, an entity model comprising, for each entity, multiple coefficients that represent at least volumetric, geometric, and material properties of the entity and multiple joints that specify joint locations and orientations of the entity. The controller 1302 outputs the entity model for subsequent use in visualization, tracking, alerting, or other entity-vision operations performed without reliance on visible-light image data.[000119] In some embodiments, the channel state information comprises a parameter selected from the group consisting of per-subcarrier values, phase, amplitude, time-of-flight, and timestamps. Associated operations may include performing a visualization operation selected from the group consisting of tracking an entity, identifying an entity, determining an alert value corresponding to an entity, and displaying an entity. The entity model may describe a posture of a corresponding entity, and an artificial intelligence component trained on labeled posture data may be utilized to describe the posture of the corresponding entity. The entity model may describe an activity of a corresponding entity, and an artificial intelligence component trained on labeled activity data may be utilized toAttorney Docket No. CURV-0003-WQdescribe the activity of the corresponding entity; in further operations, the activity of the corresponding entity may be determined based on a plurality of postures of the corresponding entity over a period of time. Identification operations may include an instance identification that maintains continuity of an entity in the specified area and an absolute identification that matches an entity using learned descriptors.[000120] The foregoing description maps to the elements shown in Fig. 13, where the controller 1302 manages inter-bank pairing and configured messaging through the antenna controllers 1306 and antenna banks 1304, the radio transceiver within the device 1300 acquires the measurements used to build the entity model, and the resulting outputs are available for downstream visualization and alerting without reliance on visible-light imagery.[000121] Again referencing Fig. 13, a wireless vision system embodied at least in part as a wireless communication device operates to detect metal entities within a specified area. A Wi-Fi vision device 1300 includes a radio transceiver in communication with a plurality of antennas arranged in at least two spatially separated banks 1304 and a controller 1302 in communication with the radio transceiver and the plurality of antennas. The radio transceiver is configured to transmit and receive non- visual frequency photons, and the controller 1302 cooperates with antenna controllers 1306 to select antenna pairs from among the plurality of antennas and to transmit, via the radio transceiver, configured messages between the selected antenna pairs over one or more non-visual frequency channels. For these configured messages, the system obtains channel state information that characterizes propagation of the non-visual frequency photons through a specified area. The controller 1302 processes the channel state information to detect, within the specified area, a metal entity by determining material-indicative features consistent with conductive or ferromagnetic objects and distinguishing the metal entity from non-metal entities. The controller 1302 then determines at least one of a presence, a location, or a relationship of the metal entity relative to another entity or structure in the specified area and outputs a detection result for subsequent use in visualization, tracking, or alerting operations performed without reliance on visible-light image data.[000122] In some embodiments, the controller 1302 further processes the channel state information to detect, within the specified area, at least one additional entity and to determine a relationship between the metal entity and the additional entity, and the detection result is output further in response to that relationship. In certain embodiments, the controller 1302 processes the channel state information to determine a posture of the additional entity and outputs the detection result further in response to the determined posture. In related embodiments, the controller 1302 processes the channel state infonuation to determine an activity of the additional entity and outputs the detection result further in response to the determined activity. The system may provide an alert in response toAttorney Docket No. CURV-0003-WQthe detection result, and may further provide a depiction of the metal entity. In some implementations, the depiction of the metal entity is provided within a rendered view of the specified area, for example via a display controller 1310 and, in some cases, coordinated with higher-level workflows hosted on an external controller 1308.[000123] Again referencing Fig. 13, a Wi-Fi vision device 1300 includes a controller 1302 configured to transmit configured messages over one or more non-visual frequency channels and to process resulting propagation measurements to create, for at least one entity within a specified area, a joint model comprising a plurality of joints that specify joint locations and orientations of the entity together with associated parameters of the joint model. Based on the parameters of the joint model, the controller 1302 determines an environment placement for the entity comprising at least a position and orientation of the entity relative to structures or other entities in the specified area, and outputs the environment placement for subsequent use in visualization, tracking, alerting, or other entity-vision operations performed without reliance on visible-light image data. The device 1300 further includes a radio transceiver in communication with a plurality of antennas arranged in at least two spatially separated banks 1304; the radio transceiver is configured to transmit and receive the non-visual frequency photons utilized in the configured messages, and the controller 1302 may cooperate with antenna controllers 1306 to select antenna pairs and schedule transmissions between the selected antenna elements. In some embodiments, the propagation data comprises channel state information for the configured messages, the channel state information including at least one parameter selected from the group consisting of per-subcarrier values, phase, amplitude, time-of-flight, and timestamps. In further embodiments, the environment placement comprises a position of the entity relative to human-use furnishings in the specified area and, in certain cases, a position of the entity relative to at least one of a seating surface or a work surface. In related embodiments, the output comprises a visualization of the specified area including the entity, and the controller 1302 applies, to the visualization, an annotation based on the environment placement.[000124] In operation as a procedure, configured messages are transmitted over one or more non-visual frequency channels, propagation data such as channel state information is obtained for those messages, and the measurements are processed to create the joint model with joint locations and orientations and associated parameters. The procedure then determines the environment placement for the entity based on the joint-model parameters and outputs the environment placement for use in downstream visualization or alerting. The procedure may further operate the radio transceiver in communication with the plurality of antennas arranged in spatially separated banks 1304, select antenna pairs from among the plurality of antennas, and transmit the configured messages between the selected antenna pairs. In some implementations, the channel state informationAttorney Docket No. CURV-0003-WQincludes at least one parameter selected from the group consisting of per-subcarrier values, phase, amplitude, time-of-flight, and timestamps.[000125] A number of procedures are described following, any one or more of which may utilize, in whole or in part, any systems, controllers, computing devices, wireless vision systems, or the like as set forth throughout the present disclosure.[000126] Referencing Fig. 14, an example procedure for generating a visualization of entities in a specified area is described following. The procedure begins by selecting antenna pairs from among a plurality of antennas that include antennas positioned in at least two spatially separated banks, corresponding to operation 1402. The procedure then transmits configured messages between the selected antenna pairs over one or more non-visual frequency channels, corresponding to operation 1404. For the configured messages, the procedure obtains channel state information that characterizes propagation of non-visual frequency photons through the specified area, corresponding to operation 1406. The channel state information is processed to infer entities within the specified area, including at least the presence and location of one or more entities, corresponding to operation 1408. Based on the processing, the procedure generates a view that is a visualization of the entities in the specified area, corresponding to operation 1410, wherein the visualization is produced without reliance on visible-light image data.[000127] In some embodiments, the at least two spatially separated banks comprise two banks of three linear antennas in each bank, and selecting antenna pairs includes pairing antennas between the two banks. In some embodiments, the banks comprise two banks of at least two omnidirectional conical antennas in each bank, and the configured messages are transmitted from and received at the omnidirectional conical antennas. In a further embodiment, the banks comprise two banks of four omnidirectional conical antennas, and selecting antenna pairs includes selecting from sixteen cross-bank antenna pairs.In certain implementations, a driving circuit operationally coupled to each omnidirectional conical antenna applies a selected polarization to increase decoupling among radiation patterns. In related embodiments, the applied polarization ensures that antennas of a given bank are separated by a polarization margin, which may include a polarization difference of at least 10 degrees.[000128] The procedure may further include transmitting the configured messages at a plurality of photon frequency ranges. In some embodiments, each of the at least two spatially separated banks is driven by an independent driving circuit. Antennas of a given bank may be separated by a polarization margin to decouple radiation patterns, and corresponding antennas of each bank may be operated in a matching polarization region to facilitate cross-bank pairing. The driving circuits mayAttorney Docket No. CURV-0003-WQcomprise solid-state circuits. In a compact deployment, the at least two spatially separated banks are housed within the same enclosure of the wireless communication device.[000129] Referencing Fig. 15, an example procedure for creating an entity model for entities within a specified area is described following. The procedure begins by selecting antenna pairs from among a plurality of antennas that include antennas positioned in at least two spatially separated banks, corresponding to operation 1402. The procedure then transmits configured messages between the selected antenna pairs over one or more non-visual frequency channels, corresponding to operation 1404. For the configured messages, the procedure obtains channel state information that characterizes propagation of non-visual frequency photons through the specified area, corresponding to operation 1406. The channel state information is processed to create, for one or more entities within the specified area, an entity model comprising, for each entity, multiple coefficients and multiple joints that specify joint locations and orientations, corresponding to operation 1508. The procedure then outputs the entity model for subsequent use in visualization, tracking, alerting, or other entity vision operations performed without reliance on visible -light image data, corresponding to operation 1510.[000130] In some embodiments, the channel state information comprises one or more parameters selected from a group such as per-subcarrier values, phase, amplitude, time-of-flight, and timestamps. In certain embodiments, the entity model further describes a posture of a corresponding entity, and an artificial intelligence component trained on labeled posture data may be utilized to generate or refine posture descriptors. In related embodiments, the entity model describes an activity of a corresponding entity, and an artificial intelligence component trained on labeled activity data may be utilized to classify activities. In further embodiments, activity determinations may be inferred from a sequence of postures for the corresponding entity over a period of time.[000131] In some implementations, visualization operations performed using the output entity model may include identifying at least one entity, where the identification may be an instance identification that maintains continuity of an entity in the specified area or an absolute identification that matches an entity using learned descriptors. Visualization operations may further include tracking one or more entities, determining an alert value in response to posture, activity, material, or volume descriptors within the entity model, and displaying one or more entities within a rendered view of the specified area.[000132] Referencing Fig. 16, an example procedure for determining a presence, location, and / or relationship of a metal entity with another entity, within a specified area, is described following. The procedure begins by selecting antenna pairs from among a plurality of antennas positioned in at least two spatially separated banks, corresponding to operation 1402. The procedure then transmitsAttorney Docket No. CURV-0003-WQconfigured messages between the selected antenna pairs over one or more non-visual frequency channels, corresponding to operation 1404. For the configured messages, the procedure obtains channel state information indicative of propagation of non-visual frequency photons through the specified area, corresponding to operation 1406. The channel state information is processed to detect one or more metal entities by determining material-indicative features consistent with conductive or ferromagnetic objects, corresponding to operation 1608. The procedure determines at least one of a presence, a location, or a relationship of the detected metal entity relative to another entity or structure in the specified area, corresponding to operation 1610. The procedure outputs a detection result in response to the determined metal entity, corresponding to operation 1612.[000133] In some embodiments, the channel state information is further processed to detect one or more additional entities within the specified area and to determine a relationship between the metal entity and the additional entities. In certain embodiments, the detection result is output further in response to the determined relationship between the metal entity and the additional entities. In some embodiments, the procedure includes determining a posture of at least one additional entity, and the detection result is output further in response to the determined posture. In certain embodiments, the procedure includes determining an activity of at least one additional entity, and the detection result is output further in response to the determined activity. In some embodiments, the procedure includes providing an alert in response to the detection result. In certain embodiments, the procedure includes providing a depiction of the detected metal entity. In further embodiments, the depiction is provided within a rendered view of the specified area.[000134] Referencing Fig. 17, an example procedure for calibrating a wireless visions system for a location of interest is described following. The procedure begins by establishing a calibration location that is similar to the location of interest, corresponding to operation 1702. A wireless vision system embodied at least in part as a wireless communication device is placed at the calibration location, corresponding to operation 1704. A controller operates to select antenna pairs and to transmit configured messages over one or more non-visual frequency channels, corresponding to operation 1706. For the configured messages, the procedure obtains channel state information indicative of propagation of non-visual frequency photons through the calibration location, corresponding to operation 1708. An artificial intelligence component processes the channel state information to calibrate parameters of the wireless vision system for detecting entities in the location of interest, including mapping wireless-vision features to entity-level detections, corresponding to operation 1710. Calibrated parameters are stored for subsequent deployment, corresponding to operation 1712.Attorney Docket No. CURV-0003-WQ[000135] In some embodiments, similarity between the calibration location and the location of interest is established within at least one similarity dimension selected from the group consisting of spatial area similarity, obstacle similarity, entity placement similarity, entity count similarity, and partition photonic response similarity. In certain embodiments, the artificial intelligence component is operated to calibrate detection of metal entities in the location of interest. In some embodiments, the artificial intelligence component processes the channel state information to create, for one or more entities, an entity model for at least one entity, the entity model comprising a plurality of coefficients that represent at least one of volumetric, geometric, or material properties of the entity and a plurality of joints that specify joint locations and orientations of the entity. In further embodiments, the artificial intelligence component processes the channel state information to determine a posture of the entity. In yet further embodiments, the artificial intelligence component processes the channel state information to determine an activity of the entity.[000136] In certain embodiments, the example procedure for calibrating a wireless vision system supports numerous benefits of the present disclosure. For example, operations utilizing non-visual photons supports detecting entities without illumination by visual or near-visual photons and enables a qualia not inherently recognizable to humans, with selective conversion to human-interpretable visualizations. The extraction and use of propagation features supports sensing through obstacles that impair visual systems and detecting large metal objects indicative of potential threats or items of interest. Calibrating with Al across defined similarity dimensions tailors performance to real environments, improving reliability in remote or lightly traveled areas, and enabling event-driven visualization that reduces privacy and compliance risks in sensitive spaces. Further, the calibration operations improve the performance of the wireless vision system at the time of installation, and reduces the time required to tune and calibrate the system for any variability present at the location of interest, and the overall cost of system installation and implementation. Implementations using standard radio hardware, solid-state circuits, multi-band operation, and compact enclosures leverage readily available power and network links, reducing installation cost, complexity, and footprint. A given procedure may support some or all of the recited benefits, and / or may support other benefits not listed.[000137] Referencing Fig. 18, an example procedure for determining joint model parameters of an entity in a specified area is described following. The procedure begins by transmitting configured messages over one or more non-visual frequency channels, corresponding to operation 1802. For the configured messages, the procedure obtains propagation data indicative of the propagation of non-visual frequency photons through the specified area, corresponding to operation 1804. The propagation data is processed to create, for at least one entity, a joint model comprising a plurality ofAttorney Docket No. CURV-0003-WQjoints that specify joint locations and orientations of the entity together with associated parameters of the joint model, corresponding to operation 1806. Based on the parameters of the joint model, the procedure determines an environment placement for the entity comprising at least a position and orientation of the entity relative to structures or other entities in the specified area, corresponding to operation 1808. The procedure then outputs the environment placement for subsequent use in visualization, tracking, alerting, or other entity vision operations performed without reliance on visible-light image data, corresponding to operation 1810.[000138] In some embodiments, the procedure operates a radio transceiver in communication with a plurality of antennas including antennas positioned in at least two spatially separated banks, selects antenna pairs from among the plurality of antennas, and transmits the configured messages between the selected antenna pairs. In certain embodiments, the propagation data comprises channel state information for the configured messages, the channel state information including at least one parameter selected from the group consisting of per-subcarrier values, phase, amplitude, time-of-flight, and timestamps.[000139] In further embodiments, the output comprises a visualization of the specified area including the entity, and an annotation is applied to the visualization based on the environment placement. In related embodiments, the environment placement comprises a position of the entity relative to human-use furnishings in the specified area, and, in some cases, relative to at least one of a seating surface or a work surface. In yet further embodiments, the annotation comprises a description of a posture of the entity or a description of an activity of the entity.[000140] Example embodiments may incorporate a Wi-Fi sensing platform that transforms Channel State Information into spatial intelligence without reliance on cameras. In such embodiments, two deep-learning components may operate in tandem: a multi-person surface and bone-joint reconstruction model that produces detailed 3D body representations (e.g., SMPL-based meshes, joint locations, detection confidence, and body-to-access-point distances) from CSI clips, and an environment surface reconstruction and semantic assignment model that generates 3D point clouds with depth and semantic labels (e.g., ground, furniture) to describe physical layouts and humanenvironment interactions. Dense CSI-derived spatial reconstructions may feed these models to output 3D coordinates and posture for people, environment mapping, and action-identification confidence ratings. Outputs may be delivered via secure data streams suitable for integration into external platforms, including text reports, 2D top-down views, and 3D rendered context views.In certain examples, training pipelines may align camera vision used for labeling with processed CSI to correlate visual appearance to CSI signatures during development while maintaining camera-free operation during deployment. To improve generalization, training may utilize diverse environmentsAttorney Docket No. CURV-0003-WQrecognizing that CSI varies with furniture, layout, and materials, enabling models to perform effectively in new spaces. Advanced reasoning layers, such as large-language-model components, may be included to interpret action sequences, infer events, and predict outcomes, supporting use cases including retail theft detection, customer behavior analytics, security threat detection, and recognition of suspicious patterns.[000141] Hardware considerations may be addressed in embodiments by standardizing antenna layouts or chipsets to reduce CSI variability across devices: alternatively, systems may accommodate heterogeneous routers through additional model training. Integrations may be facilitated through a spatial data stream API, enabling embedding into existing platforms or other Wi-Fi sensing hardware and supporting scalable, privacy-first deployments. Because the system processes Wi-Fi signals rather than images, embodiments may operate through walls or obstructions while preserving privacy and avoiding personally identifiable information, yet still provide real-time person and object detection, counts, event inteipretation, and predictive insights.[000142] Illustrative applications include school safety (e.g., detection of prohibited items or unauthorized access), retail analytics (e.g., customer interactions, layout optimization), security monitoring, elderly care fall detection, smart-home automation based on actions, manufacturing object tracking, and construction site safety. In moving beyond presence detection, embodiments may capture intention and event context by leveraging 3D spatial awareness from CSI to determine where a person or object is located and how it moves or interacts with its environment. A given system may support some or all of the recited features, and / or may support other features not listed.[000143] The methods and systems described herein may be deployed in part or in whole through a machine having a computer, computing device, processor, circuit, and / or server that executes computer readable instructions, program codes, instructions, and / or includes hardware configured to functionally execute one or more operations of the methods and systems disclosed herein. The terms computer, computing device, processor, circuit, and / or server, as utilized herein, should be understood broadly.[000144] Any one or more of the terms computer, computing device, processor, circuit, and / or server include a computer of any type, capable to access instructions stored in communication thereto such as upon a non-transient computer readable medium, whereupon the computer performs operations of systems or methods described herein upon executing the instructions. In certain embodiments, such instructions themselves comprise a computer, computing device, processor, circuit, and / or server. Additionally or alternatively, a computer, computing device, processor, circuit, and / or server may be a separate hardware device, one or more computing resources distributed across hardware devices, and / or may include such aspects as logical circuits, embedded circuits, sensors, actuators, inputAttorney Docket No. CURV-0003-WQand / or output devices, network and / or communication resources, memory resources of any type, processing resources of any type, and / or hardware devices configured to be responsive to determined conditions to functionally execute one or more operations of systems and methods herein.[000145] Network and / or communication resources include, without limitation, local area network, wide area network, wireless, internet, or any other known communication resources and protocols. Example and non-limiting hardware, computers, computing devices, processors, circuits, and / or servers include, without limitation, a general purpose computer, a server, an embedded computer, a mobile device, a virtual machine, and / or an emulated version of one or more of these. Example and non-limiting hardware, computers, computing devices, processors, circuits, and / or servers may be physical, logical, or virtual. A computer, computing device, processor, circuit, and / or server may be: a distributed resource included as an aspect of several devices; and / or included as an interoperable set of resources to perform described functions of the computer, computing device, processor, circuit, and / or server, such that the distributed resources function together to perform the operations of the computer, computing device, processor, circuit, and / or server. In certain embodiments, each computer, computing device, processor, circuit, and / or server may be on separate hardware, and / or one or more hardware devices may include aspects of more than one computer, computing device, processor, circuit, and / or server, for example as separately executable instructions stored on the hardware device, and / or as logically partitioned aspects of a set of executable instructions, with some aspects of the hardware device comprising a part of a first computer, computing device, processor, circuit, and / or server, and some aspects of the hardware device comprising a part of a second computer, computing device, processor, circuit, and / or server.[000146] A computer, computing device, processor, circuit, and / or server may be part of a server, client, network infrastructure, mobile computing platform, stationary computing platform, or other computing platform. A processor may be any kind of computational or processing device capable of executing program instructions, codes, binary instructions and the like. The processor may be or include a signal processor, digital processor, embedded processor, microprocessor or any variant such as a co-processor (math co-processor, graphic co-processor, communication co-processor and the like) and the like that may directly or indirectly facilitate execution of program code or program instructions stored thereon. In addition, the processor may enable execution of multiple programs, threads, and codes. The threads may be executed simultaneously to enhance the performance of the processor and to facilitate simultaneous operations of the application. By way of implementation, methods, program codes, program instructions and the like described herein may be implemented in one or more threads. The thread may spawn other threads that may have assigned priorities associated with them; the processor may execute these threads based on priority or any other orderAttorney Docket No. CURV-0003-WQbased on instructions provided in the program code. The processor may include memory that stores methods, codes, instructions and programs as described herein and elsewhere. The processor may access a storage medium through an interface that may store methods, codes, and instructions as described herein and elsewhere. The storage medium associated with the processor for storing methods, programs, codes, program instructions or other type of instructions capable of being executed by the computing or processing device may include but may not be limited to one or more of a CD-ROM, DVD, memory, hard disk, flash drive, RAM, ROM, cache and the like.[000147] A processor may include one or more cores that may enhance speed and performance of a multiprocessor. In embodiments, the process may be a dual core processor, quad core processors, other chip-level multiprocessor and the like that combine two or more independent cores (called a die).[000148] The methods and systems described herein may be deployed in part or in whole through a machine that executes computer readable instructions on a server, client, firewall, gateway, hub, router, or other such computer and / or networking hardware. The computer readable instructions may be associated with a server that may include a file server, print server, domain server, internet server, intranet server and other variants such as secondary server, host server, distributed server and the like. The server may include one or more of memories, processors, computer readable transitory and / or non-transitory media, storage media, ports (physical and virtual), communication devices, and interfaces capable of accessing other servers, clients, machines, and devices through a wired or a wireless medium, and the like. The methods, programs, or codes as described herein and elsewhere may be executed by the server. In addition, other devices required for execution of methods as described in this application may be considered as a part of the infrastructure associated with the server.[000149] The server may provide an interface to other devices including, without limitation, clients, other servers, printers, database servers, print servers, file servers, communication servers, distributed servers, and the like. Additionally, this coupling and / or connection may facilitate remote execution of instructions across the network. The networking of some or all of these devices may facilitate parallel processing of program code, instructions, and / or programs at one or more locations without deviating from the scope of the disclosure. In addition, all the devices attached to the server through an interface may include at least one storage medium capable of storing methods, program code, instructions, and / or programs. A central repository may provide program instructions to be executed on different devices. In this implementation, the remote repository may act as a storage medium for methods, program code, instructions, and / or programs.Attorney Docket No. CURV-0003-WQ[000150] The methods, program code, instructions, and / or programs may be associated with a client that may include a file client, print client, domain client, internet client, intranet client and other variants such as secondary client, host client, distributed client and the like. The client may include one or more of memories, processors, computer readable transitory and / or non-transitory media, storage media, ports (physical and virtual), communication devices, and interfaces capable of accessing other clients, servers, machines, and devices through a wired or a wireless medium, and the like. The methods, program code, instructions, and / or programs as described herein and elsewhere may be executed by the client. In addition, other devices utilized for execution of methods as described in this application may be considered as a part of the infrastructure associated with the client.[000151] The client may provide an interface to other devices including, without limitation, servers, other clients, printers, database servers, print servers, file servers, communication servers, distributed servers, and the like. Additionally, this coupling and / or connection may facilitate remote execution of methods, program code, instructions, and / or programs across the network. The networking of some or all of these devices may facilitate parallel processing of methods, program code, instructions, and / or programs at one or more locations without deviating from the scope of the disclosure. In addition, all the devices attached to the client through an interface may include at least one storage medium capable of storing methods, program code, instructions, and / or programs. A central repository may provide program instructions to be executed on different devices. In this implementation, the remote repository may act as a storage medium for methods, program code, instructions, and / or programs.[000152] The methods and systems described herein may be deployed in part or in whole through network infrastructures. The network infrastructure may include elements such as computing devices, servers, routers, hubs, firewalls, clients, personal computers, communication devices, routing devices and other active and passive devices, modules, and / or components as known in the art. The computing and / or non-computing device(s) associated with the network infrastructure may include, apart from other components, a storage medium such as flash memory, buffer, stack, RAM, ROM and the like. The methods, program code, instructions, and / or programs described herein and elsewhere may be executed by one or more of the network infrastructural elements.[000153] The methods, program code, instructions, and / or programs described herein and elsewhere may be implemented on a cellular network having multiple cells. The cellular network may either be frequency division multiple access (FDMA) network or code division multiple access (CDMA) network. The cellular network may include mobile devices, cell sites, base stations, repeaters, antennas, towers, and the like.Attorney Docket No. CURV-0003-WQ[000154] The methods, program code, instructions, and / or programs described herein and elsewhere may be implemented on or through mobile devices. The mobile devices may include navigation devices, cell phones, mobile phones, mobile personal digital assistants, laptops, palmtops, netbooks, pagers, electronic books readers, music players, and the like. These mobile devices may include, apart from other components, a storage medium such as a flash memory, buffer, RAM, ROM and one or more computing devices. The computing devices associated with mobile devices may be enabled to execute methods, program code, instructions, and / or programs stored thereon.Alternatively, the mobile devices may be configured to execute instructions in collaboration with other devices. The mobile devices may communicate with base stations interfaced with servers and configured to execute methods, program code, instructions, and / or programs. The mobile devices may communicate on a peer to peer network, mesh network, or other communications network. The methods, program code, instructions, and / or programs may be stored on the storage medium associated with the server and executed by a computing device embedded within the server. The base station may include a computing device and a storage medium. The storage device may store methods, program code, instructions, and / or programs executed by the computing devices associated with the base station.[000155] The methods, program code, instructions, and / or programs may be stored and / or accessed on machine readable transitory and / or non-transitory media that may include: computer components, devices, and recording media that retain digital data used for computing for some interval of time; semiconductor storage known as random access memory (RAM); mass storage typically for more permanent storage, such as optical discs, forms of magnetic storage like hard disks, tapes, drums, cards and other types; processor registers, cache memory, volatile memory, non-volatile memory; optical storage such as CD, DVD; removable media such as flash memory (e.g., USB sticks or keys), floppy disks, magnetic tape, paper tape, punch cards, standalone RAM disks, Zip drives, removable mass storage, off-line, and the like; other computer memory such as dynamic memory, static memory, read / write storage, mutable storage, read only, random access, sequential access, location addressable, file addressable, content addressable, network attached storage, storage area network, bar codes, magnetic ink, and the like.[000156] Certain operations described herein include interpreting, receiving, and / or determining one or more values, parameters, inputs, data, or other information. Operations including interpreting, receiving, and / or determining any value parameter, input, data, and / or other information include, without limitation: receiving data via a user input; receiving data over a network of any type; reading a data value from a memory location in communication with the receiving device; utilizing a default value as a received data value; estimating, calculating, or deriving a data value based on otherAttorney Docket No. CURV-0003-WQinformation available to the receiving device; and / or updating any of these in response to a later received data value. In certain embodiments, a data value may be received by a first operation, and later updated by a second operation, as part of the receiving a data value. For example, when communications are down, intermittent, or interrupted, a first operation to interpret, receive, and / or determine a data value may be performed, and when communications are restored an updated operation to interpret, receive, and / or determine the data value may be performed.[000157] Certain logical groupings of operations herein, for example methods or procedures of the current disclosure, are provided to illustrate aspects of the present disclosure. Operations described herein are schematically described and / or depicted, and operations may be combined, divided, reordered, added, or removed in a manner consistent with the disclosure herein. It is understood that the context of an operational description may require an ordering for one or more operations, and / or an order for one or more operations may be explicitly disclosed, but the order of operations should be understood broadly, where any equivalent grouping of operations to provide an equivalent outcome of operations is specifically contemplated herein. For example, if a value is used in one operational step, the determining of the value may be required before that operational step in certain contexts (e.g. where the time delay of data for an operation to achieve a certain effect is important), but may not be required before that operation step in other contexts (e.g. where usage of the value from a previous execution cycle of the operations would be sufficient for those purposes). Accordingly, in certain embodiments an order of operations and grouping of operations as described is explicitly contemplated herein, and in certain embodiments re-ordering, subdivision, and / or different grouping of operations is explicitly contemplated herein.[000158] The methods and systems described herein may transform physical and / or or intangible items from one state to another. The methods and systems described herein may also transform data representing physical and / or intangible items from one state to another.[000159] The elements described and depicted herein, including in flow charts, block diagrams, and / or operational descriptions, depict and / or describe specific example arrangements of elements for purposes of illustration. However, the depicted and / or described elements, the functions thereof, and / or arrangements of these, may be implemented on machines, such as through computer executable transitory and / or non-transitory media having a processor capable of executing program instructions stored thereon, and / or as logical circuits or hardware arrangements. Example arrangements of programming instructions include at least: monolithic structure of instructions; standalone modules of instructions for elements or portions thereof; and / or as modules of instructions that employ external routines, code, services, and so forth; and / or any combination of these, and all such implementations are contemplated to be within the scope of embodiments of theAttorney Docket No. CURV-0003-WQpresent disclosure Examples of such machines include, without limitation, personal digital assistants, laptops, personal computers, mobile phones, other handheld computing devices, medical equipment, wired or wireless communication devices, transducers, chips, calculators, satellites, tablet PCs, electronic books, gadgets, electronic devices, devices having artificial intelligence, computing devices, networking equipment, servers, routers and the like. Furthermore, the elements described and / or depicted herein, and / or any other logical components, may be implemented on a machine capable of executing program instructions. Thus, while the foregoing flow charts, block diagrams, and / or operational descriptions set forth functional aspects of the disclosed systems, any arrangement of program instructions implementing these functional aspects are contemplated herein. Similarly, it will be appreciated that the various steps identified and described above may be varied, and that the order of steps may be adapted to particular applications of the techniques disclosed herein.Additionally, any steps or operations may be divided and / or combined in any manner providing similar functionality to the described operations. All such variations and modifications are contemplated in the present disclosure. The methods and / or processes described above, and steps thereof, may be implemented in hardware, program code, instructions, and / or programs or any combination of hardware and methods, program code, instructions, and / or programs suitable for a particular application. Example hardware includes a dedicated computing device or specific computing device, a particular aspect or component of a specific computing device, and / or an arrangement of hardware components and / or logical circuits to perform one or more of the operations of a method and / or system. The processes may be implemented in one or more microprocessors, microcontrollers, embedded microcontrollers, programmable digital signal processors or other programmable device, along with internal and / or external memory. The processes may also, or instead, be embodied in an application specific integrated circuit, a programmable gate array, programmable array logic, or any other device or combination of devices that may be configured to process electronic signals. It will further be appreciated that one or more of the processes may be realized as a computer executable code capable of being executed on a machine readable medium.[000160] The computer executable code may be created using a structured programming language such as C, an object oriented programming language such as C++, or any other high-level or low-level programming language (including assembly languages, hardware description languages, and database programming languages and technologies) that may be stored, compiled or interpreted to run on one of the above devices, as well as heterogeneous combinations of processors, processor architectures, or combinations of different hardware and computer readable instructions, or any other machine capable of executing program instructions.Attorney Docket No. CURV-0003-WQ[000161] Thus, in one aspect, each method described above and combinations thereof may be embodied in computer executable code that, when executing on one or more computing devices, performs the steps thereof. In another aspect, the methods may be embodied in systems that perform the steps thereof, and may be distributed across devices in a number of ways, or all of the functionality may be integrated into a dedicated, standalone device or other hardware. In another aspect, the means for performing the steps associated with the processes described above may include any of the hardware and / or computer-readable instructions described above. All such permutations and combinations are contemplated in embodiments of the present disclosure.[000162] While the disclosure has been disclosed in connection with the preferred embodiments shown and described in detail, various modifications and improvements thereon will become readily apparent to those skilled in the art. Accordingly, the spirit and scope of the present disclosure is not to be limited by the foregoing examples, but is to be understood in the broadest sense allowable by law.

Claims

Attorney Docket No. CURV-0003-WQWhat is claimed is:

1. A wireless vision system comprising a wireless communication device, the system comprising:a radio transceiver and a plurality of antennas including antennas positioned in at least two spatially separated banks, wherein the radio transceiver is configured to transmit and receive nonvisual frequency photons;a controller in communication with the radio transceiver and the plurality of antennas, the controller configured to:select antenna pairs from among the plurality of antennas;transmit, via the radio transceiver, configured messages between selected antennas of the wireless communication device over one or more non- visual frequency channels;obtain, for the configured messages, channel state information that characterizes propagation of the non-visual frequency photons through a specified area;process the channel state information to infer one or more entities within the specified area, including at least presence and location of at least one of the one or more entities; and generate a view that is a visualization of the at least one of the one or more entities in the specified area based on the channel state information.

2. The wireless vision system of claim 1, wherein the channel state information comprises complex propagation descriptors including at least one of: per-subcarrier values, phase, amplitude, time-of-flight, or timestamps of the configured messages.

3. The wireless vision system of claim 1 , wherein the at least two spatially separated banks comprise at least one arrangement selected from:two banks of three linear antennae in each bank; ortwo banks of at least two omnidirectional conical antennae in each bank.

4. The wireless vision system of claim 1, wherein the at least two spatially separated banks comprise two banks of four omnidirectional conical antennae.

5. The wireless vision system of claim 1, wherein the at least two spatially separated banks comprise two banks of at least two omnidirectional conical antennae in each bank, wherein each one of the at least two omnidirectional conical antennae are operationally coupled to a driving circuit configured to apply a selected polarization to each of the at least two omnidirectional conical antennae in each bank.

6. The wireless vision system of claim 5, wherein the driving circuit is further configured to apply the selected polarization to ensure each antenna of the at least two omnidirectional conical antennae are separated by a polarization margin.Attorney Docket No. CURV-0003-WQ7. The wireless vision system of claim 6, wherein the polarization margin comprises a polarization difference of at least 10 degrees.

8. The wireless vision system of claim 1, wherein the controller is configured to transmit the configured messages at a plurality of photon frequency ranges.

9. The wireless vision system of claim 8, wherein the plurality of photon frequency ranges span at least the radio and microwave spectrum.

10. The wireless vision system of claim 8, wherein the plurality of photon frequency ranges comprise a plurality of frequency values between 1 MHz and 1 GHz, inclusive.

11. The wireless vision system of claim 8, wherein the plurality of photon frequency ranges comprise a plurality of frequency values between 104Hz and 1010Hz.

12. The wireless vision system of claim 8, wherein the plurality of photon frequency ranges comprise a plurality of frequency values between KT and 109Hz.

13. The wireless vision system of claim 1, wherein each of the at least two spatially separated banks are driven by independent driving circuits.

14. The wireless vision system of claim 13, wherein antennae of a given spatially separated bank are separated by a polarization margin.

15. The wireless vision system of claim 14, wherein corresponding antennae of each spatially separated bank are configured to operate in a matching polarization region.

16. The wireless vision system of claim 14, wherein the driving circuits comprise solid state circuits.

17. The wireless vision system of claim 1, wherein the at least two spatially separated banks are present in a same housing of the wireless vision system.

18. A method of wireless vision, comprising:selecting antenna pairs from among a plurality of antennas including antennas in at least two spatially separated banks;transmitting, via a radio transceiver, configured messages between the selected antenna pairs of a wireless communication device over one or more non- visual frequency channels;obtaining, for the configured messages, channel state information that characterizes propagation of non-visual frequency photons through a specified area;processing the channel state information to infer one or more entities within the specified area, including at least presence and location of at least one of the one or more entities; and generating a view that is a visualization of the entities in the specified area based on the channel state information, the view being produced without reliance on visible-light image data.Attorney Docket No. CURV-0003-WQ19. The method of claim 18. wherein the at least two spatially separated banks comprise two banks of three linear antennas in each bank, and selecting antenna pairs comprises pairing antennas between the two banks.

20. The method of claim 18, wherein the at least two spatially separated banks comprise two banks of at least two omnidirectional conical antennas in each bank, and transmitting the configured messages comprises transmitting from and receiving at the omnidirectional conical antennas.

21. The method of claim 20, wherein the at least two spatially separated banks comprise two banks of four omnidirectional conical antennas, and selecting antenna pairs comprises selecting from sixteen cross-bank antenna pairs.

22. The method of claim 20, further comprising applying, via a driving circuit operationally coupled to each omnidirectional conical antenna, a selected polarization to each of the omnidirectional conical antennas to increase decoupling among radiation patterns.

23. The method of claim 22, wherein applying the selected polarization comprises ensuring that antennas of a given bank are separated by a polarization margin.

24. The method of claim 23, wherein the polarization margin comprises a polarization difference of at least 10 degrees.

25. The method of claim 18, further comprising transmitting the configured messages at a plurality of photon frequency ranges.

26. The method of claim 18, wherein each of the at least two spatially separated banks is driven by an independent driving circuit.

27. The method of claim 26, wherein antennas of a given spatially separated bank are separated by a polarization margin to decouple radiation patterns.

28. The method of claim 27, wherein corresponding antennas of each spatially separated bank are operated in a matching polarization region to facilitate cross-bank pairing.

29. The method of claim 26, wherein the driving circuits comprise solid-state circuits.

30. The method of claim 18, further comprising housing the at least two spatially separated banks within a same enclosure of the wireless communication device.

31. The method of claim 18, wherein the channel state information comprises complex propagation descriptors including at least one of: per-subcarrier values, phase, amplitude, time-of-flight, or timestamps of the configured messages.

32. A wireless vision system comprising a wireless communication device for creating a Wi-Fi vision based entity model, the system comprising:a radio transceiver and a plurality of antennas including antennas positioned in at least two spatially separated banks;Attorney Docket No. CURV-0003-WQa controller in communication with the radio transceiver and the plurality of antennas, wherein the radio transceiver is configured to transmit and receive non-visual frequency photons; wherein the controller is configured to:select antenna pairs from among the plurality of antennas; transmit, via the radio transceiver, configured messages between the selected antenna pairs over one or more non-visual frequency channels;obtain, for the configured messages, channel state information that characterizes propagation of the non-visual frequency photons through a specified area;process the channel state information to create, for one or more entities within the specified area, an entity model comprising, for each of the one or more entities:a plurality coefficients that represent at least one of volumetric, geometric, or material properties of the entity; anda plurality of joints that specify joint locations and orientations of the entity; andoutput the entity model for subsequent use in a visualization operation.

33. The system of claim 32, wherein the visualization operations comprises at least one operation selected from: tracking an entity of the one or more entities; identifying an entity of the one or more entities; determining an alert value corresponding to at least one of the one or more entities; or displaying at least one of the one or more entities.

34. The system of claim 32, wherein the channel state information comprises at least one parameter selected from the group consisting of: per-subcarrier values, phase, amplitude, time-of-flight, and timestamps.

35. The system of claim 32, wherein the entity model describes a posture of the corresponding entity.

36. The system of claim 35, wherein the controller is further configured to implement an artificial intelligence component trained on labeled posture data for entities to describe the posture of the corresponding entity.

37. The system of claim 32, wherein the entity model describes an activity of the corresponding entity.

38. The system of claim 37, wherein the controller is further configured to implement an artificial intelligence component trained on labeled activity data for entities to describe the activity of the corresponding entity.Attorney Docket No. CURV-0003-WQ39. The system of claim 37, wherein the controller is further configured to determine the activity of the corresponding entity based on a plurality of postures of the corresponding entity over a period of time.

40. The system of claim 38, wherein the controller is further configured to determine the activity of the corresponding entity based on a plurality of postures of the corresponding entity over a period of time.

41. The system of claim 32, wherein the visualization operation comprises identifying at least one entity of the one or more entities.

42. The system of claim 41, wherein the identifying comprises an instance identification.

43. The system of claim 41, wherein the identifying comprises an absolute identification.

44. A method of wireless vision, comprising:selecting antenna pairs from among a plurality of antennas including antennas in at least two spatially separated banks;transmitting, via a radio transceiver, configured messages between the selected antenna pairs over one or more non-visual frequency channels;obtaining, for the configured messages, channel state information that characterizes propagation of non-visual frequency photons through a specified area;processing the channel state information to create, for one or more entities within the specified area, an entity model comprising, for each entity:multiple coefficients that represent at least one of volumetric, geometric, or material properties of the entity; andmultiple joints that specify joint locations and orientations of the entity; andoutputting the entity model for subsequent use in visualization, tracking, alerting, or other entity vision operations performed without reliance on visible-light image data.

45. The method of claim 44, wherein the channel state information comprises at least one parameter selected from the group consisting of per-subcarrier values, phase, amplitude, time-of-flight, and timestamps.

46. The method of claim 44, wherein the entity model describes a posture of a corresponding entity.

47. The method of claim 46, further comprising implementing an artificial intelligence component trained on labeled posture data for entities to describe the posture of the corresponding entity.Attorney Docket No. CURV-0003-WQ48. The method of claim 44. wherein the entity model describes an activity of a corresponding entity.

49. The method of claim 48, further comprising implementing an artificial intelligence component trained on labeled activity data for entities to describe the activity of the corresponding entity.

50. The method of claim 48, further comprising determining the activity of the corresponding entity based on a plurality of postures of the corresponding entity over a period of time.

51. The method of claim 49, further comprising determining the activity of the corresponding entity based on a plurality of postures of the corresponding entity over a period of time.

52. The method of claim 44, wherein the visualization operation comprises identifying at least one entity of the one or more entities.

53. The method of claim 52, wherein identifying comprises an instance identification.

54. The method of claim 52, wherein identifying comprises an absolute identification.

55. The method of claim 44, wherein performing the visualization operation comprises tracking at least one entity of the one or more entities.

56. The method of claim 44, wherein performing the visualization operation comprises determining an alert value corresponding to at least one entity of the one or more entities.

57. The method of claim 56, wherein determining the alert value comprises determining the alert value in response to, for the at least one entity of the one or more entities, at least one of: a posture of the entity, an activity of the entity, a material of the entity, or a volume of the entity.

58. The method of claim 44. wherein performing the visualization operation comprises displaying at least one entity of the plurality of entities.

59. A wireless vision system comprising a wireless communication device for creating a Wi-Fi vision scheme, comprising:a radio transceiver and a plurality of antennas including antennas positioned in at least two spatially separated banks;a controller in communication with the radio transceiver and the plurality of antennas, wherein the radio transceiver is configured to transmit and receive non-visual frequency photons, and wherein the controller is configured to:select antenna pairs from among the plurality of antennas; transmit, via the radio transceiver, configured messages between the selected antenna pairs over one or more non-visual frequency channels;Attorney Docket No. CURV-0003-WQobtain, for the configured messages, channel state information that characterizes propagation of the non-visual frequency photons through a specified area;process the channel state information to detect, within the specified area, a metal entity by determining material-indicative features consistent with conductive or ferromagnetic objects;determine at least one of a presence, a location, or a relationship of the metal entity relative to another entity or structure in the specified area; and output a detection result in response to the determined metal entity.

60. The system of claim 59, wherein the controller is further configured to process the channel state information to detect, within the specified area, at least one additional entity, and to determine a relationship between the metal entity and the at least one additional entity.

61. The system of claim 59, wherein the controller is further configured to output the detection result further in response to the relationship between the metal entity and the at least one additional entity.

62. The system of claim 60, wherein the controller is further configured to process the channel state information to determine a posture of the at least one additional entity.

63. The system of claim 62, wherein the controller is further configured to output the detection result further in response to the posture of the at least one additional entity.

64. The system of claim 60, wherein the controller is further configured to process the channel state information to determine an activity of the at least one additional entity.

65. The system of claim 64, wherein the controller is further configured to output the detection result further in response to the activity of the at least one additional entity.

66. The system of claim 59, wherein the controller is further configured to provide an alert in response to the detection result.

67. The system of claim 66, wherein the controller is further configured to provide a depiction of the metal entity.

68. The system of claim 59, wherein the controller is further configured to provide a depiction of the metal entity.

69. A method of implementing a wireless vision scheme, the method comprising: selecting antenna pairs from among a plurality of antennas including antennas in at least two spatially separated banks;transmitting, via a radio transceiver, configured messages between the selected antenna pairs over one or more non-visual frequency channels;Attorney Docket No. CURV-0003-WQobtaining, for the configured messages, channel state information that characterizes propagation of non-visual frequency photons through a specified area;processing the channel state information to detect, within the specified area, a metal entity by determining material-indicative features consistent with conductive or ferromagnetic objects;determining at least one of a presence, a location, or a relationship of the metal entity relative to another entity or structure in the specified area; andoutputting a detection result in response to the determined metal entity.

70. The method of claim 69, further comprising processing the channel state information to detect, within the specified area, at least one additional entity and determining a relationship between the metal entity and the at least one additional entity.

71. The method of claim 70, further comprising outputting the detection result further in response to the relationship between the metal entity and the at least one additional entity.

72. The method of claim 70, further comprising determining a posture of the at least one additional entity.

73. The method of claim 72, further comprising outputting the detection result further in response to the posture of the at least one additional entity.

74. The method of claim 70, further comprising determining an activity of the at least one additional entity.

75. The method of claim 74, further comprising outputting the detection result further in response to the activity of the at least one additional entity.

76. The method of claim 69, further comprising providing an alert in response to the detection result.

77. The method of claim 69, further comprising providing a depiction of the metal entity.

78. The method of claim 77, further comprising providing a depiction of the metal entity within the specified area.

79. A method of calibrating a wireless vision system for a location of interest, the method comprising:establishing a calibration location that is similar to the location of interest within at least one similarity dimension;placing, at the calibration location, a wireless vision system comprising a wireless communication device having a radio transceiver, a plurality of antennas including antennas in at least two spatially separated banks, and a controller;Attorney Docket No. CURV-0003-WQoperating the controller to select antenna pairs from among the plurality of antennas and to transmit, via the radio transceiver, configured messages between the selected antenna pairs over one or more non-visual frequency channels;obtaining, for the configured messages, channel state information that characterizes propagation of non-visual frequency photons through the calibration location;operating an artificial intelligence component to process the channel state information to calibrate parameters of the wireless vision system for detecting entities in the location of interest, including at least mapping wireless vision features derived from the channel state information to entity-level detections; andstoring the calibrated parameters for subsequent deployment of the wireless vision system at the location of interest.

80. The method of claim 79, wherein the similarity dimension comprises at least one feature similarity selected from the group consisting of: spatial area similarity, obstacle similarity, entity placement similarity, entity count similarity, and partition photonic response similarity.

81. The method of claim 79, further comprising operating the artificial intelligence component for detecting metal entities in the location of interest.

82. The method of claim 79, further comprising operating the artificial intelligence component to process the channel state information to create, for one or more entities within the location of interest, an entity model for at least one entity of the one or more entities.

83. The method of claim 82, wherein the entity model comprises a plurality of coefficients that represent at least one of volumetric, geometric, or material properties of the entity.

84. The method of claim 83, wherein the entity model further comprises a plurality of joints that specify joint locations and orientations of the entity.

85. The method of claim 79, further comprising operating the artificial intelligence component to process the channel state information to determine a posture of the entity.

86. The method of claim 79, further comprising operating the artificial intelligence component to process the channel state information to determine an activity of the entity.

87. A wireless vision system comprising:a wireless communication device;a controller configured to:transmit configured messages over one or more non-visual frequency channels;obtain, for the configured messages, propagation data indicative of propagation of non-visual frequency photons through a specified area;Attorney Docket No. CURV-0003-WQprocess the propagation data to create, for at least one entity within the specified area, a joint model comprising a plurality of joints that specify joint locations and orientations of the entity and associated parameters of the joint model;determine, based on the parameters of the joint model, an environment placement for the entity comprising at least a position and orientation of the entity relative to structures or other entities in the specified area; and output the environment placement for subsequent use in visualization, tracking, alerting, or other entity vision operations performed without reliance on visible-light image data.

88. The system of claim 87, further comprising:a plurality of antennas including antennas positioned in at least two spatially separated banks and a radio transceiver in communication with the controller and the plurality of antennas;wherein the radio transceiver is configured to transmit and receive the non- visual frequency photons utilized in the configured messages; andwherein the controller is further configured to select antenna pairs from among the plurality of antennas and to transmit the configured messages between the selected antenna pairs to generate the propagation data.

89. The system of claim 88, wherein the propagation data comprises channel state information for the configured messages.

90. The system of claim 89, wherein the channel state information comprises at least one parameter selected from the group consisting of: per-subcarrier values, phase, amplitude, time-of-flight, and timestamps.

91. The system of claim 87, wherein the environment placement comprises a position of the entity relative to human-use furnishings in the specified area.

92. The system of claim 87, wherein the environment placement comprises a position of the entity relative to at least one of a seating surface or a work surface in the specified area.

93. The system of claim 87, wherein the output comprises a visualization of the specified area including the entity, and wherein the controller is further configured to apply, to the visualization, an annotation based on the environment placement.

94. The system of claim 93, wherein the annotation comprises a description of a posture of the entity.

95. The system of claim 93, wherein the annotation comprises a description of an activity of the entity.Attorney Docket No. CURV-0003-WQ96. A method of wireless vision embodied at least in part by a wireless communication device, the method comprising:transmitting configured messages over one or more non-visual frequency channels; obtaining, for the configured messages, propagation data indicative of propagation of non-visual frequency photons through a specified area;processing the propagation data to create, for at least one entity within the specified area, a joint model comprising a plurality of joints that specify joint locations and orientations of the entity and associated parameters of the joint model;determining, based on the parameters of the joint model, an environment placement for the entity comprising at least a position and orientation of the entity relative to structures or other entities in the specified area; andoutputting the environment placement for subsequent use in visualization, tracking, alerting, or other entity vision operations performed without reliance on visible-light image data.

97. The method of claim 96, further comprising:operating a radio transceiver in communication with a plurality of antennas including antennas positioned in at least two spatially separated banks;selecting antenna pairs from among the plurality of antennas; andtransmitting the configured messages between the selected antenna pairs.

98. The method of claim 96, wherein the propagation data comprises channel state information for the configured messages, the channel state information comprising at least one parameter selected from the group consisting of per-subcarrier values, phase, amplitude, time-of-flight, and timestamps.

99. The method of claim 96, wherein the output comprises a visualization of the specified area including the entity, the method further comprising applying, to the visualization, an annotation based on the environment placement.

100. The method of claim 96, wherein the environment placement comprises a position of the entity relative to human-use furnishings in the specified area.

101. The method of claim 96, wherein the environment placement comprises a position of the entity relative to at least one of a seating surface or a work surface in the specified area.

102. The method of claim 99, wherein the annotation comprises a description of a posture of the entity.

103. The method of claim 99, wherein the annotation comprises a description of an activity of the entity.

Images