3754 results about "Recognition algorithm" patented technology

A gaze tracking technique is implemented with a head mounted gaze tracking device that communicates with a server. The server receives scene images from the head mounted gaze tracking device which captures external scenes viewed by a user wearing the head mounted device. The server also receives gaze direction information from the head mounted gaze tracking device. The gaze direction information indicates where in the external scenes the user was gazing when viewing the external scenes. An image recognition algorithm is executed on the scene images to identify items within the external scenes viewed by the user. A gazing log tracking the identified items viewed by the user is generated.

Systems and methods for two-way, interactive communication regarding emergency notifications and responses for mobile environments are disclosed. In accordance with one embodiment of the present invention, a specific geographic area is designated for selective emergency communications. The emergency communications may comprise text, audio, video, and other types of data. The emergency notification is sent to users' mobile communications devices such as in-vehicle telematics units, cellular phones, personal digital assistants (PDAs), laptops, etc. that are currently located in the designated area. The sender of the emergency message or the users' service provider(s) may remotely control cameras and microphones associated with the users' mobile communications devices. For example, a rear camera ordinarily used when driving in reverse may be used to capture images and video that may assist authorities in searching for a suspect. The users' vehicles may send photographs or video streams of nearby individuals, cars and license plates, along with real-time location information, in response to the emergency notification. Image recognition algorithms may be used to analyze license plates, vehicles, and faces captured by the users' cameras and determine whether they match a suspect's description. Advantageously, the present invention utilizes dormant resources in a highly beneficial and time-saving manner that increases public safety and national security.

Methods involving extraction of information from the inherent variability of physiometrics, including data on cardiovascular and pulmonary functions such as heart rate variability, characteristics of ECG traces, pulse, oxygenation of subcutaneous blood, respiration rate, temperature or CO₂ content of exhaled air, heart sounds, and body resonance, can be used to identify individual subjects, particularly humans. Biometric data for use in the methods can be obtained either from contact sensors or at a distance. The methods can be performed alone or can be fused with previous identification algorithms.

A phishing detection agent is provided. In one embodiment, a user's browser includes a plug-in application or agent that may capture a visual record of a webpage and, with a cached copy of known, authentic websites provided to it via periodic updates, perform a series of image comparison functions to determine if the suspected website is attempting to deceive the user. The phishing detection agent is capable of performing an image recognition algorithm, such as logo recognition algorithm, optical character recognition, an image similarity algorithm, or combination of two or more of the above. If the suspected webpage corresponds to one of the authentic web pages, but the domain name of the suspected web page does not match the domain name of one of the authentic web pages, the suspected web page is flagged as a phishing web site.

The personalized content system of the system combines a summary music identification value creation and identification algorithm that represents a mathematical summary music identification of a song, an audio file, or other relational music criteria and data (e.g. title, artist, genre, style, beats per minute, etc.) or any combination thereof. The derived value represents the musical taste or style attributes of a song or style. The analysis and generation of the summary music identification value is intended to take place while outside the vehicle where attention can be paid and vehicle distraction and safety is not an issue. A user can generate a single summary music identification value or can generate multiple summary music identification values for different criteria. In other cases, summary music identification values can be defined for genre, but with a much more personal touch possible than with fixed playlist radio stations. In another embodiment, a user can obtain and use someone else's summary music identification value (e.g. a celebrity, artist, friend, etc.) and use it with the system.

Systems and methods for recognizing the gestures of an entity, notably a human being, and, optionally, for controlling an electrical or electronic system or apparatus, are discussed. The system uses sensors that measure signals, preferentially representative of inertial data about the movements of said entity, and implements a process for enriching a dictionary of said gestures to be recognized and a recognition algorithm, for recognition among the classes of gestures in said dictionary. The algorithm implemented is preferably of the dynamic time warping type. The system carries out preprocessing operations, such as the elimination of signals captured during periods of inactivity of the entity, subsampling of the signals, and normalization of the measurements by reduction, and preferentially uses, to classify the gestures, specific distance calculation modes and modes for merging or voting between the various measurements by the sensors.

The invention discloses a cross-camera pedestrian detection tracking method based on depth learning, which comprises the steps of: by training a pedestrian detection network, carrying out pedestrian detection on an input monitoring video sequence; initializing tracking targets by a target box obtained by pedestrian detection, extracting shallow layer features and deep layer features of a region corresponding to a candidate box in the pedestrian detection network, and implementing tracking; when the targets disappear, carrying out pedestrian re-identification which comprises the process of: after target disappearance information is obtained, finding images with the highest matching degrees with the disappearing targets from candidate images obtained by the pedestrian detection network and continuously tracking; and when tracking is ended, outputting motion tracks of the pedestrian targets under multiple cameras. The features extracted by the method can overcome influence of illuminationvariations and viewing angle variations; moreover, for both the tracking and pedestrian re-identification parts, the features are extracted from the pedestrian detection network; pedestrian detection, multi-target tracking and pedestrian re-identification are organically fused; and accurate cross-camera pedestrian detection and tracking in a large-range scene are implemented.

Providing enhanced video content includes processing at least one video feed through at least one spatiotemporal pattern recognition algorithm that uses machine learning to develop an understanding of a plurality of events and to determine at least one event type for each of the plurality of events. The event type includes an entry in a relationship library detailing a relationship between two visible features. Extracting and indexing a plurality of video cuts from the video feed is performed based on the at least one event type determined by the understanding that corresponds to an event in the plurality of events detectable in the video cuts. Lastly, automatically and under computer control, an enhanced video content data structure is generated using the extracted plurality of video cuts based on the indexing of the extracted plurality of video cuts.

An input system for mobile search and a method therefor are provided. The input system includes an input module receiving a code input for a specific term and a voice input corresponding thereto, a database including a glossary and an acoustic model, wherein the glossary includes a plurality of terms and a sequence list, and each of the terms has a search weight based on an order of the sequence list, a process module selecting a first number of candidate terms from the glossary according to the code input by using an input algorithm and obtaining a second number of candidate terms by using a speech recognition algorithm to compare the voice input with the first number of candidate terms via the acoustic model, wherein the second number of candidate terms are listed in a particular order based on their respective search weights, and an output module showing the second number of candidate terms in the particular order for selecting the specific term therefrom.

A system and method of audio processing provides enhanced speech recognition. Audio input is received at a plurality of microphones. The multi-channel audio signal from the microphones may be processed by a beamforming network to generate a single-channel enhanced audio signal, on which voice activity is detected. Audio signals from the microphones are additionally processed by an adaptable noise cancellation filter having variable filter coefficients to generate a noise-suppressed audio signal. The variable filter coefficients are updated during periods of voice inactivity. A speech recognition engine may apply a speech recognition algorithm to the noise-suppressed audio signal and generate an appropriate output. The operation of the speech recognition engine and the adaptable noise cancellation filter may advantageously be controlled based on voice activity detected in the single-channel enhanced audio signal from the beamforming network.

A processor integrating and controlling at least two A/V devices by constructing a control model, referred to as a filter graph, of the at least two A/V devices as a function of a physical connection topology of the at least two A/V devices and a desired content to be rendered by one of the at least two A/V devices. The filter graph may be constructed as a function of at least two device filters corresponding to the at least two A/V devices, in which the device filters include certain characteristics of the at least two A/V device. These characteristics may include the input or output pins for each device, the media type that the A/V device may process, the type of functions that the device may serve, etc. The desired content may be received as a user input which is entered via a keyboard, mouse or other comparable input devices. In addition, the user input may be entered as a voice command, which may be parsed by the processor using conventional speech recognition algorithms or natural language processing to extract the necessary information. Once the filter graph is constructed, the processor may control the at least two A/V devices via the filter graph by invoking predetermined operations on the filter graph resulting in the appropriate commands being sent to the at least two A/V devices, thereby results in the rendering of the desired content.

The present invention relates to methods for evaluating the probability that a patient's diagnosis may be treated with a particular clinical regimen or therapy.

Embodiments of the invention perform assisted tagging of images, including tagging of people, locations, and activities depicted in those images. A batch of images is received comprising images of faces, including at least some faces that have not yet been tagged. A facial recognition algorithm is applied to the faces to determine matching data comprising possible tags for each untagged face. A logic engine applies logic rules to reduce the likelihood that certain matches are correct. The most likely match from among the possible matches is selected for suggestion to the user for verification. Once verified, the metadata of the image indicating the recognized people within the image is updated.

The invention discloses a named entities recognition method based on bidirectional LSTM and CRF. The named entities recognition method based on the bidirectional LSTM and CRF is improved and optimizedbased on the traditional named entities recognition algorithm in the prior art. The named entities recognition method based on the bidirectional LSTM and CRF comprises the following steps: (1) preprocessing a text, extracting phrase information and character information of the text; (2) coding the text character information by means of the bidirectional LSTM neural network to convert the text character information into character vectors; (3) using the glove model to code the text phrase information into word vectors; (4) combining the character vectors and the word vectors into a context information vector and putting the context information vector into the bidirectional LSTM neural network; and (5) decoding the output of the bidirectional LSTM with a linear chain condition random field to obtain a text annotation entity. The invention uses a deep neural network to extract text features and decodes the textual features with the condition random field, therefore, the text feature information can be effectively extracted and good effects can be achieved in the entity recognition tasks of different languages.

A system for authentication and/or authorization which incorporates two communication channels, and at least one of third-party data sources, geographic correlation algorithms, speech recognition algorithms, voice biometric comparison algorithms, and mechanisms to convert textual data into speech. A site visitor's identity can be verified using one or all of such features in combination with a visitors address on one of the channels.

A method for searching a collection of digital images on a display screen, comprising: entering an image review mode and displaying on the display screen a first digital image from the collection of digital images; designating a face contained in the first digital image by using an interactive user interface to indicate a region of the displayed first digital image containing the face; using a processor to execute an automatic face recognition algorithm to identify one or more additional digital images from the collection of digital images that contain the designated face; and displaying the identified one or more additional digital images on the display screen.

The present invention discloses a 3D (three-dimensional) convolutional neural network based human body behavior recognition method, which is mainly used for solving the problem of recognition of a specific human body behavior in the fields of computer vision and pattern recognition. The implementation steps of the method are as follows: (1) carrying out video input; (2) carrying out preprocessing to obtain a training sample set and a test sample set; (3) constructing a 3D convolutional neural network; (4) extracting a feature vector; (5) performing classification training; and (6) outputting a test result. According to the 3D convolutional neural network based human body behavior recognition method disclosed by the present invention, human body detection and movement estimation are implemented by using an optical flow method, and a moving object can be detected without knowing any information of a scenario. The method has more significant performance when an input of a network is a multi-dimensional image, and enables an image to be directly used as the input of the network, so that a complex feature extraction and data reconstruction process in a conventional recognition algorithm is avoided, and recognition of a human body behavior is more accurate.

A MMR system for newspaper publishing comprises a plurality of mobile devices, an MMR gateway, an MMR matching unit and an MMR publisher. The MMR matching unit receives an image query from the MMR gateway and sends it to one or more of the recognition units to identify a result including a document, the page and the location on the page. The MMR system also includes a quality predictor as a plug-in installed on the mobile device to filter images before they are included as part of a retrieval request or as part of the MMR matching unit. The quality predictor comprises an input for receiving recognition algorithm information, a vector calculator, a score generator and a scoring module. The quality predictor receives as inputs an image query, context information and device parameters, and generates an outputs a recognizability score. The present invention also includes a method for generating robustness features.

The present invention discloses a parallelized human body behavior identification method. According to the method, skeleton data of Kinect is used as input; a distributed behavior identification algorithm is implemented based on a Spark computing framework; and a complete parallel identifying process is formed. Acquisition of the skeleton data of a human body is based on scene depth acquisition capacity of Kinect and the data is preprocessed to ensure invariability of displacement and scale of characteristics; and a human body structural vector, joint included angle information and skeleton weight bias are respectively selected for static behavior characteristics and a dynamic behavior searching algorithm for a structural similarity is provided. On the identification algorithm, a neural network algorithm is parallelized on Spark; a quasi-newton method L-BFGS is adopted to optimize a network weight updating process; and the training speed is obviously increased. According to an identification platform, a Hadoop distributed file system HDFS is used as a behavior data storage layer; Spark is applied to a universal resource manager YARN; the parallel neural network algorithm is used as an upper application; and the integral system architecture has excellent extendibility.