512 results about "Object detection" patented technology

Provided herein are system, apparatus, device, method and / or computer program product embodiments, and / or combinations and sub-combinations thereof, for improving object detection based on gaze direction. In the method, a first image of surroundings of a vehicle is received. A second image of a driver of the vehicle is also received. The second image captured contemporaneously with the first image. Based on the second image, a gaze direction of the driver is determined. Based on the gaze direction, a region of the first image being viewed by the driver is determined. An object detection algorithm is applied to recognize an object in the surroundings of the vehicle at a greater level of detail within the region.

The application relates to the technical field of microfluidic chip, and discloses a paper chip presetting method. The method comprises the following steps: cooling a paper chip to a temperature of-60 to-10 DEG C; and adding reagents dropwise to the cooled paper chip to perform pre-freezing and freeze-drying. The paper chip presetting method improves the preset amount and uniformity of reagents on the paper chip, expands the upper limit of target object detection of the paper chip, and improves the detection accuracy.

This specification provides an object detection method and a data storage unit detection method. The object detection method includes: performing anomaly detection on multiple network addresses to determine abnormal network addresses, wherein any one of the multiple network addresses exists in at least two network address sets corresponding to target objects, and each target object corresponds to a network address set; determining an abnormal network address set based on the abnormal network addresses and the association between the multiple network addresses and the network address sets, and identifying the target objects corresponding to the abnormal network address sets as abnormal objects; thereby accurately and quickly locating abnormal objects that have suffered network attacks, avoiding the problem of not being able to accurately identify abnormal objects that have suffered network attacks.

Systems and methods are provided for long-tailed distribution object detection that utilizes a streamlined framework using multi-stage supervised and / or semi-supervised training. The disclosed technology eliminates reliance on extensive labeled datasets by effectively utilizing unlabeled data through pseudo-labeling and robust augmentation techniques. The disclosed technology enhances detection accuracy across both frequent and rare categories without introducing unnecessary complexities such as knowledge distillation or meta-learning. Certain implementations of the disclosed technology include multi-stage training with external unlabeled images. Certain implementations of the disclosed technology include pre-training a model on the (frequent) head classes and learning is transferred to the (rare) tail classes, which provides significant benchmarking improvements over other approaches.

Provided are an image analysis system and method. The image analysis system comprises: a local device that generates, as one or more types of base data, detection results acquired for locally collected images by using an object detection model; and an inference device that includes one or more inference modules and receives the one or more types of base data from the local device, wherein the inference device distributes the base data to the one or more inference modules according to the type of base data, and collects, for each predetermined identifier, one or more inference results inferred from each of the inference modules to generate inference information for each identifier.

When retraining an AI model used for object detection processing on edge devices to add identifiable classes, the aim is to improve the class identification accuracy of the AI ​​model while reducing the workload on the user. [Solution] The information processing device comprises an annotation processing unit that performs annotation processing on an input image using a large-scale AI model that performs object detection processing and is configured to output region information indicating an object detection region even for objects that cannot be classified as a class; a reception processing unit that accepts corrections from the user regarding the annotation results from the annotation processing unit; a large-scale model retraining processing unit that performs retraining processing on the large-scale AI model using the annotation information corrected by the user; and an edge model retraining processing unit that performs retraining processing on an edge model, which is an AI model used in an edge device that performs object detection processing on captured images, by knowledge distillation using the retrained large-scale AI model as the training model.

The present invention relates to the technical field of object detection. Disclosed are a training method and apparatus for a railway foreign object phrase localization model, and a device and a medium. Cross alignment can be performed on an image feature vector and a text feature vector, so as to obtain an aligned image feature vector and an aligned text feature vector; the pair of the aligned image feature vector and the aligned text feature vector is then used to perform training, such that a model can accurately align a foreign object described in a text with a target region in an image, thereby ensuring accurate phrase localization in a complex scenario, and being applicable to fine-grained scenario description in railway foreign object detection; and a trained railway foreign object phrase localization model can thus effectively execute and complete railway foreign object phrase localization tasks, thereby effectively implementing phrase localization of railway foreign objects.

The application discloses a control method and device of an access gate, a storage medium and an electronic device. The method comprises: acquiring a time sequence image sequence corresponding to a passing area, wherein the passing area comprises an access gate, and the time sequence image sequence comprises a plurality of continuous image frames; performing an object detection operation on each image frame in the time sequence image sequence to determine object information of a target object; in the case that the object information meets object recognition conditions, performing a state prediction operation on the target object to generate a gate control signal matched with a passing action of the target object; and adjusting a motion trajectory parameter of the access gate according to the gate control signal, wherein the motion trajectory parameter comprises at least one of an opening angle and a rotating speed. The application solves the technical problem of low passing efficiency caused by the rigid opening and closing actions of the gate and the response lag of the gate when the passing speed of the user is inconsistent and the body size difference is large.

To provide a technology that improves learning accuracy of a neural network model that detects an object from an image.SOLUTION: A learning device acquires a degree of similarity between a correct answer region indicating a region of an object in an image and each of a plurality of anchor boxes preset in the image, and selects an anchor box with a degree of similarity equal to or greater than a predetermined threshold among the plurality of anchor boxes for the correct answer region. The learning device learns, on the basis of the correct answer region and the selected anchor box, a neural network model for object detection. The learning device changes, when anchor boxes of the upper limit number are selected for the correct answer region, on the basis of the degree of similarity obtained for the anchor boxes selected for the correct answer region, the upper limit number of the anchor boxes for the correct answer region.SELECTED DRAWING: Figure 1

This invention proposes a method and system for detecting missing weld points on vehicle bodies based on 3D vision and global matching, belonging to the field of image recognition technology. The method includes: determining the transformation relationship between the camera coordinate system and the vehicle body coordinate system through calibration; at each observation position, a 3D camera acquires an image I containing a frame of 2D texture and the corresponding 3D data; inputting the acquired image I into a pre-trained deep learning object detection model, outputting the bounding box information (u, v, w, h) for each weld point; for each detected weld point, calculating the set S of the weld point's 3D coordinates in the vehicle body coordinate system. det For the theoretical solder joint library S cad Each theoretical solder joint P in cad,i In set S det The invention employs a "detect first, match later" strategy, which does not rely on the precise projection of the theoretical position, thus reducing the requirements for the positioning accuracy of the production line tooling fixtures and conveying system.

This application discloses an open-world 3D target detection method and apparatus, comprising: acquiring multimodal data of a target scene; performing 3D target detection on 3D point cloud data to obtain 3D candidate targets; inputting the geometric features of each 3D candidate target into an out-of-distribution target classifier to determine whether the 3D candidate target belongs to a known category set and to identify unknown targets; projecting the 3D candidate targets corresponding to the unknown targets onto 2D image data, obtaining the 2D image region, and inputting it into a visual language model, using natural language prompts to guide the visual language model to output the semantic category name of the unknown targets; fusing the semantic category name and the spatial location information of the 3D candidate targets to generate open-world 3D detection results. This application can effectively identify unknown 3D targets in an open world and generate open-world 3D detection results that combine spatial positioning and semantic description.

A method includes receiving a new image of an item based on a time-stamped user interaction with the item, applying object detection to the new image to identify the item, determining a respective similarity of the image to each old image of the item in a library of old images, and adding the new image to the library only if each respective similarity is below a similarity threshold.

This invention discloses a method, device, and storage medium for identifying potential hazards in power transmission channels based on collaborative reasoning using object detection and knowledge graphs. The method generates a knowledge graph of potential hazards in power transmission channels, a rule base for reasoning about potential hazards in power transmission channels, and an object detection model. The object detection model detects and identifies multiple current entity target categories from the current power transmission channel image. These multiple current entity target categories are associated with entity types in the knowledge graph of potential hazards in power transmission channels, resulting in a knowledge graph of potential hazards in power transmission channels that incorporates a transient semantic subgraph. Then, the knowledge graph of potential hazards in power transmission channels that incorporates a transient semantic subgraph is traversed and matched based on the rule base for reasoning about potential hazards in power transmission channels. Reasoning is performed using the obtained hazard reasoning rules to obtain the hazard identification result for the current power transmission channel. The device and storage medium are used to implement the method. This invention can obtain accurate hazard reasoning results, and the hazard logical reasoning process is interpretable.

To enable accurate detection of the subject being tracked. [Solution] An information processing device (100) for detecting a subject to be tracked from a video image captured by an imaging device (200) and performing automatic tracking photography, comprising: an obstacle detection means (153) for detecting obstacles that may be mistaken for the subject from the video image by AI object detection; a registration means (154) for registering the video region of the obstacle detected by the obstacle detection means (153); an exclusion means (155) for excluding the video region of the obstacle registered by the registration means (154) from a detection region for detecting the subject in the video image; and a subject detection means (156) for detecting the subject from the detection region excluding the video region of the obstacle by AI object detection.

This invention discloses a method for detecting corn kernel breakage rate by incorporating an attention mechanism. This method further improves the lightweight one-stage object detection network YOLOv4-tiny by introducing a coordinate attention mechanism into the feature pyramid network to enhance the model's ability to detect dense corn kernels. Furthermore, an online mosaic data augmentation method is used during the model training phase to improve the model's ability to detect smaller-scale broken kernels. The method disclosed in this invention is easier to deploy on embedded platforms due to its low parameter size, and it has a high response speed while maintaining accuracy. It can efficiently detect the degree of corn kernel breakage and thus obtain the corn kernel breakage rate in real time.

According to the present invention, activities such as monitoring and surveying using multiple mobile units can be carried out more efficiently. [Solution] A control system that controls the operation of multiple mobile bodies equipped with measurement sensors capable of detecting objects in order to detect the object, comprising: an object detection unit that detects the object based on measurement data acquired by the measurement sensors; and a unit formation command determination unit that determines a formation command specifying the formation configuration of one or more platoons to be formed by the multiple mobile bodies, wherein the unit formation command determination unit determines the formation command of the platoon according to the current state or future predicted state of the detected object when the object detection unit detects the object.

The disclosure relates to system and method for continuously tracking humans in an area. The method includes receiving video data of the area from overhead cameras. Each of overhead cameras includes Field of View (FoV), FoV includes overlapping region and non-overlapping region, and overlapping region corresponds to region of intersection between at least two FoVs. The method further includes detecting presence humans in first FoV through object detection and classification models; for each human of humans, assigning unique global identity (ID) corresponding to human in first FoV, and reassigning unique global ID to human when human moves from first FoV to second FoV through overlapping region between first FoV and second FoV using weighted combination of resource assignment algorithm, intersection-over-union (IOU) based track detection, and velocity and direction estimation of subsequent frame of video data; and continuously tracking, in real-time, each of humans in the area through unique global ID.

A method and an electronic device for generating one or more sequences of media are provided. The method includes obtaining ongoing media currently played on the electronic device. The method includes selecting a transition object as a RoI in the ongoing media based on a user input or an object detection. The method includes generating follow-up media based on a description of the transition object. The follow-up media indicates a content sequence aligned with spatial-temporal attributes of the transition object. The method includes generating a transformation effect based on motion information. The transformation effect indicates a transition of objects present in an outgoing frame of the ongoing media to the objects present in an incoming frame of the follow-up media. The method includes displaying the follow-up media with the transformation effect, thereby generating the one or more sequences of the media.

This application discloses a method for detecting apple disease spores using microscopic data based on multimodal and semi-supervised learning. The method includes: acquiring raw microscopic data of apple disease fungal spores, performing edge detection and texture enhancement to generate texture-enhanced data; extracting features from the raw microscopic data and texture-enhanced data using a dual-branch encoder, and fusing them using a cross-attention mechanism to obtain enhanced visual features; inputting the textual description information of the disease fungal spores into a text encoder for encoding to obtain global text features; aligning the enhanced visual features and global text features across modalities based on a multimodal object detection network, outputting multimodal fused features, and inputting them into a semi-supervised learning framework to train a student-teacher model using labeled and unlabeled data; inputting the microscopic data to be detected into the trained model and outputting the detection results. This method improves the detection accuracy and robustness of microscopic data with extremely low annotation costs.

A computer-implemented method for detecting objects within an advanced driver assistance system (ADAS) is provided. The method includes obtaining road scene datasets from a plurality of cameras, including at least road scene images and road scene data annotations, to be provided to an object detection neural network communicating with an open-vocabulary detector of a vehicle, converting, by a text prompter, the road scene data annotations into natural text inputs, converting, by a text embedder, the natural text inputs into embeddings, minimizing objective functions during training to adjust parameters of the object detection neural network, and detecting, by the object detection neural network, objects within the road scene datasets to provide alerts or notifications to a driver of the vehicle pertaining to the detected objects.

To effectively prevent shoplifting at self-checkout counters. [Solution] Server 1 acquires object detection information indicating the detection result of a predetermined object that has entered a predetermined area A by the object detection sensor S1. Server 1 acquires scanned information indicating that the product information of a predetermined product B has been read by the barcode scanner BS. If one of the object detection information and the scanned information is acquired but the other is not, Server 1 determines that there was an abnormality in the payment transaction by the prospective product purchaser U, i.e., there is a possibility that the prospective product purchaser U has shoplifted.

The application discloses a radar object detection method and device and a storage medium, and belongs to the technical field of radars. The method comprises the following steps: acquiring a depth feature map sent by at least one radar node; mapping the depth feature map to a world coordinate system according to preset spatial alignment parameters, to generate a perspective feature map; weighting and fusing the perspective feature map through an attention mechanism according to spatial position information of the radar node and the perspective feature, to generate a fused feature map; and performing target detection based on the fused feature map, to generate a three-dimensional air situation. Through radar networking and power adjustment, the detection effect of the radar is improved through the cooperative detection capability of low-cost devices.

The present disclosure provides an object detection method and related equipment. The method comprises: receiving three-dimensional data and image data for object detection; calculating a first feature of the three-dimensional data and a second feature of the image data; calculating a corresponding first object query feature based on the three-dimensional data; calculating a corresponding second object query feature based on the three-dimensional data and a projection relationship between the three-dimensional data and the image data; calculating attention based on the first feature, the second feature, the first object query feature and the second object query feature to obtain attention data; and outputting an object detection result based on the attention data.

An object recognition system includes object detection circuitry configured to detect two or more portions of an object as a detection target captured in a frame image input from a camera, fitness calculation circuitry configured to calculate fitness as a recognition target of an object as a detection target based on positions and sizes of the two or more portions, comparison circuitry configured to compare the fitness as the recognition target with a predetermined reference value, and object recognition circuitry configured to recognize only an object as the detection target that has cleared the reference value as a result of the comparison.

The invention relates to a method that comprises determining (501) a first object detection result. If the first object detection result satisfies each activation condition of a driving assistance function, an activation signal is transmitted (504) to a control device that obtains (506) a light intensity map corresponding to a light pattern and indicative of light intensity values for controlling light elements of a pixelated light source of a light module. The light intensity map is transmitted (507) for projecting (508) a first pixelated light beam according to the light pattern. After projection, a first image representative of the scene is obtained (509) and a depth map is determined (510). A second object detection result is determined (511) from the determined depth map.

The application relates to a reinforcement learning optimization method and device for an end-to-end target detection algorithm. The method comprises the following steps: obtaining image data to be processed, extracting image data features of the image data to be processed by using a hyperparameter optimization model, selecting corresponding hyperparameters in combination with corresponding task information, performing algorithm optimization on an end-to-end algorithm by using the hyperparameters, obtaining an optimized end-to-end algorithm, performing target recognition by using the optimized end-to-end algorithm, evaluating a target detection recognition result, selecting corresponding rewards, updating the hyperparameter optimization model based on the rewards in combination with a reinforcement learning algorithm gradient, and obtaining an optimized parameter optimization model. The method of the application does not need manual design for optimization, does not need to set a fixed manual detection threshold, greatly improves the algorithm optimization efficiency, and significantly improves the generalization of the algorithm.

A method, a device and a non-transitory computer-readable storage medium for setting a confidence threshold for objects detected in a region of a captured scene based on historical confidence scores for objects detected in that region.

The application relates to the field of airport pavement nondestructive testing and artificial intelligence technology, in particular to an abnormal body detection model and method for airport management large model knowledge base construction. The model comprises a multi-scale local structure parser, which is used for extracting multiple spatial resolution local textures and geometric features from ground penetrating radar images; a cross-domain semantic benchmark unit, which comprises an image branch of a pre-trained visual-linguistic joint encoder and is used for mapping images of different airports to their benchmark semantic fields, generating common semantic benchmark features decoupled from the airport background; a feature fusion unit, which is used for interactive fusion of local features and benchmark features to generate enhanced joint feature representation; and an abnormal body positioning and classification head, which is used for determining the position and category of the underground abnormal body. The application improves the generalization ability and accuracy of the model in detecting underground abnormal bodies of unknown airports.