136 results about "Object detector" patented technology

Techniques for determining a presence of an object, especially an object such as animal or debris, in a path of a vehicle, are discussed herein. For example, sensors of various modalities, which may include multispectral sensors, may capture data representing an environment the vehicle is traversing. In examples, one or more trained machine learned (ML) models, operating on a vehicle computing system, may detect and / or classify objects in the environment, based on input data of one or more modalities or spectral bands. The ML models may be pre-trained using training data including real sensor data, synthetic data, and / or augmented data, along with auto-generated annotations. In some examples, hyperspectral data may be used to identify materials associated with detected objects. A confidence score associated with the detection of the object may also be computed. The vehicle may be controlled based on detection of the object and its classification.

Techniques are discussed herein for generating three-dimensional (3D) representations of an environment based on two-dimensional (2D) image data, and using the 3D representations to perform 3D object detection and other 3D analyses of the environment. 2D image data may be received, along with depth estimation data associated with the 2D image data. Using the 2D image data and associated depth data, an image-based object detector may generate 3D representations, including point clouds and / or 3D pixel grids, for the 2D image or particular regions of interest. In some examples, a 3D point cloud may be generated by projecting pixels from the 2D image into 3D space followed by a trained 3D convolutional neural network (CNN) performing object detection. Additionally or alternatively, a top-down view of a 3D pixel grid representation may be used to perform object detection using 2D convolutions.

The invention belongs to the field of computer vision and artificial intelligence security, and discloses a self-adaptive frequency domain adversarial training method and device for a target detector, the target detector comprises a repair module and a pedestrian detector which are connected in series, and the input of the repair module is connected with the output of a patch detector; the self-adaptive frequency domain adversarial training method comprises the following steps: losses in joint training comprise standard target detection losses, repair consistency losses on a frequency domain based on a frequency domain image corresponding to a training image and a clean image, and repair dependence losses based on a detected average precision mean value; according to the invention, the end-to-end joint training is carried out through the restoration module and the subsequent pedestrian detector, and the optimization target of the restoration module is directly aligned with the improvement of the detection robustness, so that the confrontation disturbance is eliminated as far as possible, and meanwhile, the key semantic information of the detection task is reserved to the maximum extent. The separation of the performance of the repair module and the pedestrian detector is avoided, and the detection robustness is improved.

Systems and methods for a multi-camera object detector having two or more backbone models. In particular, systems and methods are provided for including two or more backbone machine learning models, with one backbone optimized for speed and the other backbone optimized for precision. In particular, the first backbone can be highly accurate but slower than the second backbone and with a higher computer resource usage. The second backbone can be fast and efficient but have lower accuracy for object detection. In some examples, the second backbone can use fewer images and / or lower resolution images. The determination of which backbone to use can be based on fixed rules, or it can be determined based on another machine learning component. The outputs from the first and second backbones for each camera can be combined together into a unified representation, such as a bird's eye view (BEV) space.

Systems, methods, and storage media for manifesting a virtual object in a virtual environment are disclosed. Exemplary embodiments may: receive, at a first physical object detector, a first signal, from a first physical object-associated element in a first physical environment; identify, at a first value identification module, based on the first signal, a first value associated with the first signal; identify, at a first virtual object identification module, based on the first value, a first virtual object; and manifest, at a first virtual environment output device, a first manifestation of the first virtual object in a first manifestation of the first virtual environment.

An auto-exposure control is proposed for high dynamic range images, along with a neural network for exposure selection that is trained jointly, end-to-end with an object detector and an image signal processing (ISP) pipeline. Corresponding method and system for high dynamic range object detection are also provided.

An auto-exposure control is proposed for high dynamic range images, along with a neural network for exposure selection that is trained jointly, end-to-end with an object detector and an image signal processing (ISP) pipeline. Corresponding method and system for high dynamic range object detection are also provided.

An auto-exposure control is proposed for high dynamic range images, along with a neural network for exposure selection that is trained jointly, end-to-end with an object detector and an image signal processing (ISP) pipeline. Corresponding method and system for high dynamic range object detection are also provided.

3D objection detection is a computer vision task that generally refers to detecting (e.g. classifying and localizing) an object in 3D space from an image or video that captures the object. This computer vision task has many useful applications, such as autonomous driving applications which rely on the detection of 3D objects in a local environment to make autonomous driving decisions. State-of-the-art 3D object detectors generally rely on machine learning, but current training processes for these detectors do not specifically address false negative detections, or missed objects, which are often caused by occlusions and / or cluttered backgrounds in the given image / video. Reducing false negatives is crucial for many downstream applications, particularly autonomous driving applications which rely on accurate detection of obstacles for making safe driving decisions. The present disclosure provides for a multi-stage training process that reduces false negative detections by 3D object detectors.

A vehicle blind-spot reduction device includes side-view mirrors on sides of a vehicle, an actuator configured to move an optical display of each of the side-view mirrors, a detector configured to detect a display position of the optical display, a surrounding environment information acquirer configured to acquire environment information on surrounding of the vehicle, a moving object detector configured to detect a moving object running beside the vehicle on the basis of the surrounding environment information, a blind spot setter configured to set a blind spot of each of the side-view mirrors on the basis of the display position of the optical display, and a display surface adjuster. Based on determining that the moving object has entered the blind spot, the display surface adjuster causes the actuator to operate to adjust the display position of the optical display.

Systems and methods for site-based calibration of object detection rules, such as for surveillance video cameras, are described. Video data from a video image sensor may be processed using an object detector to determine object data for a detected object. The object data may be post-processed using a post-processing rule set to determine whether the detected object violates the post-processing rule set. Event notifications to a video surveillance application may be prevented responsive to the object data violating the post-processing rule set.

An auto-exposure control is proposed for high dynamic range images, along with a neural network for exposure selection that is trained jointly, end-to-end with an object detector and an image signal processing (ISP) pipeline. Corresponding method and system for high dynamic range object detection are also provided.

An image capturing apparatus includes an image capturing unit, a driving unit configured to drive the image capturing unit, and at least one processor. The at least one processor is configured to function as a motion vector detector configured to detect motion vectors based on image data output from the image capturing unit, an object detector configured to detect a plurality of moving objects based on the motion vectors, and a controlling unit configured to perform tracking control by controlling the driving unit. The controlling unit calculates respective evaluation values of the plurality of moving objects based on at least one of (a) information on the plurality of moving objects, (b) information on a shake of the image capturing apparatus, and (c) information on a driving state of the driving unit. The controlling unit controls the driving unit based on the evaluation values.

Provided is a training method for updating a deep learning-based object detector of an autonomous vehicle to adapt to a driving environment, the method comprising (a)(i) inputting a training image corresponding to the driving environment into an environment-specific object detector to cause the kth environment-specific object detector to (i-1) generate an environment-specific feature map through convolution, (i-2) generate an environment-specific pooling feature map through ROI pooling, and (i-3) generate environment-specific object detection information by applying a full connection operation to the kth environment-specific pooling feature map, (ii) inputting the environment-specific feature map into an environment-specific ranking network, (ii-1) generating an environment-specific segmentation map through an environment-specific deconvolution layer, and (ii-2) generating an environment-specific ranking score through an environment-specific discriminator, and (b) training the environment-specific object detector, training the environment-specific deconvolution layer, and training the environment-specific discriminator.

A headlamp device includes a headlamp, an object detector, and a controller. The headlamp is mounted on a vehicle and includes a light emitting unit composed of multiple light emitting cells. The object detector detects an object around the vehicle and generates an object detection signal including object coordinates corresponding to a location of the detected object. The controller controls light of the headlamp based on information on the object coordinates. The controller may control the light emitting unit such that at least some light emitting cells are turned off according to an object detection signal. Each of multiple light emitting units includes a circuit board, the multiple light emitting cells separated from each other on the circuit board, a molding member formed between the light emitting structures, and a protective member formed on the molding member to surround sides of the multiple wavelength conversion members while filling a gap between the multiple wavelength conversion members.

A driving assistance device includes: a vehicle surrounding image aquisitor acquiring a captured image of surroundings of a vehicle or by processing the captured image, as a vehicle surrounding image where the surroundings of the vehicle are visually recognized from a set viewpoint; an object detector detecting an object of attention being an object present around the vehicle requiring attention; a model image acquisitor acquiring a model image indicating an appearance of the object of attention; an assistance image generator generating an assistance image where the model image is synthesized with the vehicle surrounding image according to a position of the object of attention; and an image display displaying the assistance image on a display device such that the vehicle surrounding image excluding the model image has a luminance difference or a contrast ratio with the model image increasing stepwise as a distance from the viewpoint increases.

A power transmitter (101) provides wireless power to a power receiver (105). The power transmitter (101) comprises a transmitter coil (103) generating a power transfer signal. A communicator (307) is arranged to communicate using a communication carrier and an object detector (is arranged to detect objects. An update circuit (315) receives (701) a power receiver identity modulated on communication carrier by a power receiver in response a detection of the power receiver. The identity is stored and transmitted to an update device by modulation of the communication carrier when the update device is detected. Update data modulated on the communication carrier by the update device is received and stored with a link to the identity. When a power receiver is subsequently detected, a new identity is received (711) and update data linked to this identity is retrieved (713). The retrieved update data is then transmitted to the power receiver which may then perform an update using the update data.

Systems and techniques are described for suppressing false positive notifications for detected objects. In various examples, first bounding box data indicating a detection of a first class of object in a first frame of image data may be received from an object detector. Second bounding box data indicating a prior detection of the first class of object in a second frame of image data may be determined. A first value representing a similarity between the first bounding box data and the second bounding box data may be determined. A notification associated with the detection of the first class of object in the first frame of image data may be suppressed based at least in part on the first value.

The present application relates to a kind of method and system for detecting general object based on dynamic inference network.The method comprises: based on the given general object detector, construct the dynamic inference object detector of multiple outlet;For the dynamic inference object detector of multiple outlet, insert multi-scale adaptive gating network;Using the training strategy of no hyperparameter, train the dynamic inference object detector of multiple outlet and the multi-scale adaptive gating network;Using the dynamic inference object detector of multiple outlet and the multi-scale adaptive gating network of training completion, using variable time delay inference strategy carries out general object detection.The present application can be widely applied in the deployment of a variety of general object detector, and then be applied to intelligent security, automatic driving, unmanned aerial vehicle survey and a variety of application scenarios, to realize the effective deployment of the same model in a variety of hardware requirements, application, reduce the consumption of manpower, material resources and financial resources.

A host satellite comprises a first memory segment for storing a mission operation lookup table having at least one mission operation identifier associated with mission parameter constraints, a target object type and object detection model parameters associated with the target object type. A first sensor provides satellite sensor data that represents orbital mission characteristics. A second sensor captures images. A mission operation selector is responsive to the orbital mission characteristics and the mission parameter constraints for selecting the at least one mission operation identifier. An object detector is responsive to the captured images and the object detection model parameters associated with the selected mission operation identifier for detecting objects from the captured images. An analytics generator is responsive to the detected objects, the orbital mission characteristics, and the mission operation identifier for creating a mission analytics packet. A communication interface transfers the mission analytics packet to the client terminal.

The present invention relates to a computer-implemented method of training an object detector (OD), an object detector (OD), a computer program and a computer-readable (storage) medium. In order to carry out the method, a set of object classes (C) obfuscated by a neural network (NN), an object feature map (FM) originating from the neural network (NN), and object class tags (LO) comprising object classes (C) assigned to each object (O) in the object feature map (FM) need to be acquired. Subsequently, a training head (TH) configured to determine the authenticity of an object class (C) assigned to the object (O) in the object feature map (FM) from the object feature map (FM) and the object class tag (LO) is added to the neural network (NN). A modified object category tag (LO ') is determined from the set of object categories (C) and the object category tag (LO), wherein at least one object category assignment (C) is modified. The modified object category tags (LO ') are processed together with the object feature map (FM) by a neural network, generating an output, which is compared with an authenticity tag (LT) using a first objective function (OF1), the authenticity tag comprising an indication of the authenticity of each object category (C) in the modified object category tags (LO'). And updating the network parameters according to the comparison result.

Methods and systems for automated double emulsion droplet library generators are described herein. Optimizing the geometry and composition of double emulsion droplets is essential for maximizing their utility in various scientific and industrial applications. The microfluidic method is effective to generate a droplet library, which is useful to optimize droplets. However, monitoring and intervention by a human are necessary to generate droplet library. The disclosure provided herein presents an automated double-droplet library generator that utilizes a fine-tuned object detector and feedback control. The system autonomously produces libraries of single-core double emulsion droplets tailored to user-defined sizes, shell thicknesses, and solute concentrations.

This invention presents an active learning method based on uncertainty and diversity acquisition functions, belonging to the field of active learning for object detectors in computer vision. It addresses the problem that existing active learning processes, utilizing initialized object detectors, cannot deeply mine valuable image features. The method includes: using a weakly supervised object detector trained on image-level labels and a fully supervised object detector initialized with partial instance-level labels to detect images in the training set that have been labeled with ground truth or pseudo-ground truth values; calculating instance-level difficulty scores, class-level difficulty scores, and class similarity weights based on the detection results; then detecting images without instance-level labels, calculating image-level difficulty scores based on class relevance weights and entropy, and determining candidate images; calculating the similarity between every two candidate images based on the detection results, determining cluster centers for the candidate images, and identifying valuable images after clustering. This invention further refines fully supervised object detectors.

Embodiments can relate to a computer vision object detection system having a processor with object detector (OD) and a light informed shape analysis (LISA) modules. Upon receiving an image scene, the process can cause the OD module to: to identify an object as an object of interest; identify a region of interest for the object of interest; generate a bounding box encompassing the object of interest and the region of interest; and track movement of the object of interest. When the OD detector drops the bounding box for the object of interest, the processor can cause the LISA module to: extract a region of interest for use as an expected region of interest for the object of interest; extract one or more shape contours of the object of interest for use as a representation of the object of interest; and track movement of the object of interest.

The application discloses a kind of dynamic visual monitoring method and device for high-altitude falling object, mainly by high-altitude falling object detector and trajectory predictor two parts consist of, cover two functions of high-altitude falling object event detection based on event camera, high-altitude falling object trajectory prediction, can be realized in certain complex environment for the accurate identification and trajectory tracking of falling object target.Because the event camera itself high dynamic range, low delay etc., compared with the monitoring method based on traditional optical camera, the present application can be applicable to more complex real environment and bring higher monitoring ability.Project results can provide a new solution for the current high-altitude falling object monitoring supervision problem, and bring greater social and economic benefits to society, have important practical significance.

This application describes a method for automatically labeling data samples with uncertainty quantification. The proposed method comprises the steps of simultaneously feeding an input data sample into an online object detection module with uncertainty estimation and an off-board object detector with uncertainty estimation; comparing the resulting data from both the online classification and uncertainty estimation module and the off-board classification and uncertainty estimation module; and deciding whether to discard the input data sample, send it for human labeling, or archive it along with the classification and uncertainty estimation from the off-board object detection module with uncertainty estimation.These uncertainty estimates are then used in the training phase of subsequent object detectors to weight different samples differently, with human-labeled samples and automatically labeled samples with high confidence receiving more weight.

The RetinAI invention provides a novel integrated system designed for home-based early screening and detection of ocular tumors and retinal diseases, leveraging advanced artificial intelligence and optical imaging technology. The system comprises a wearable headset with precisely aligned imaging sensors and dual-mode illumination sources for capturing extraocular images to detect leukocoria and intraocular retinal images for direct visualization of tumors, eliminating the need for pharmacological pupil dilation. On-device deep learning models, including representative object detectors and convolutional neural networks, provide real-time diagnostic outputs. Clinical validations demonstrate RetinAI's capability to detect leukocoria as small as 1 mm in diameter and identify a broad spectrum of retinal diseases, confirming its significant potential to enhance ocular health screening, particularly in underserved and resource-limited settings. The cost-effective, user-friendly, and portable nature of the RetinAI system positions it as an essential tool for improving access to ophthalmic care globally.

The invention discloses a bidirectional tracking truth value generation method and system based on MLLM and 3D perception, and the method comprises the steps: S1, inputting image data and 3D laser point cloud data which are synchronous in time sequence, processing each frame of point cloud through a 3D target detector, obtaining a 3D detection frame, and then projecting the 3D detection frame to a synchronous image, and obtaining a corresponding 2D BBOX; s2, performing forward tracking and reverse tracking on the 3D detection frame of each frame to generate a forward track and a reverse track, then performing track merging on the forward track and the reverse track, and identifying a suspicious track in the merging process; s3, for the suspicious trajectory, extracting corresponding images and 2D BBOX, constructing textualized and structured semantic descriptions for each group of images and 2D BBOX data, inputting the textualized and structured semantic descriptions to a pre-trained MLLM, outputting semantic association scores and an inference chain by the MLLM, and performing association decision on the suspicious trajectory based on the semantic association scores; and S4, outputting time sequence truth value data including the complete object ID, the 3D detection frame of each frame and the object category.

Systems, devices, and methods related to an object detector and a Deep learning accelerator are described. For example, a computing apparatus has an integrated circuit device with the Deep learning accelerator configured to execute instructions generated by a compiler from a description of an artificial neural network of the object detector. The artificial neural network includes a first cross stage partial network to extract features from an image and a second cross stage partial network to combine the features to identify a region of interest in the image showing an object. The artificial neural network uses a technique of minimum cost assignment in assigning a classification to the object and thus avoids post processing of non-maximum suppression.