A moving state analysis device improves accuracy of moving state recognition by including a detection unit configured to detect, from image data associated with a frame, an object and a region of the object, for each of frames that constitute first video data captured in a course of movement of a first moving body, and a learning unit configured to learn a DNN model that takes video data and sensor data as input and that outputs a probability of each moving state, based on the first video data, a feature of first sensor data measured in relation to the first moving body and corresponding to a capture of the first video data, a detection result of the object and the region of the object, and information that indicates a moving state associated with the first video data.

A feature point detection apparatus for detecting feature points in image data includes an image data providing unit for providing the image data, a key point determination unit for determining key points in the image data, a feature determination unit for determining features associated with the key points, each describing a local environment of a key point in the image data, and a feature point providing unit for providing the feature points. A feature point is represented by the position of a key point in the image data and the associated features. The image data comprise intensity data and associated depth data, and the determination of the key points and the associated features is based on a local analysis of the image data in dependence on both the intensity data and the depth data.

The present disclosure provides an image processing method. An image to be classified is input into a feature extraction model to generate N dimensional features. Dimension fusion is performed on M features of the N dimensional features to obtain M dimension fusion features. The image to be classified is processed based on M dimension fusion features and remaining features of the N dimensional features other than the M features.

Embodiments of the present disclosure disclose a method, apparatus, and device for fusing features applied to small target detection, and a storage medium, relate to the field of computer vision technology. A particular embodiment of the method for fusing features applied to small target detection comprises: acquiring feature maps output by convolutional layers in a Backbone network; performing convolution on the feature maps to obtain input feature maps of feature layers, the feature layers representing resolutions of the input feature maps; and fusing, based on densely connection feature pyramid network features, the input feature maps of each feature layer to obtain output feature maps of the feature layer. Since no additional convolutional layer is introduced for feature fusion, the detection performance for small targets may be enhanced without additional parameters, and the detection ability for small targets may be improved with computing resource constraints.

This disclosure includes technologies for action recognition in general. The disclosed system may automatically detect various types of actions in a video, including reportable actions that cause shrinkage in a practical application for loss prevention in the retail industry. Further, appropriate responses may be invoked if a reportable action is recognized. In some embodiments, a three-branch architecture may be used in a machine learning model for action and/or activity recognition. The three-branch architecture may include a main branch for action recognition, an auxiliary branch for learning/identifying an actor (e.g., human parsing) related to an action, and an auxiliary branch for learning/identifying a scene related to an action. In this three-branch architecture, the knowledge of the actor and the scene may be integrated in two different levels for action and/or activity recognition.

A method, device and computer program product, the method comprising: obtaining access to a classifier trained upon a multiplicity of sets of decoded coefficients; obtaining a set of block coefficients associated with at least a part of the compressed image; and applying the classifier to the set of block coefficients, to obtain a classification of the compressed image.

Provided is a system and method for fusion recognition using an active stick filter. The system for fusion recognition using the active stick filter includes a data input unit configured to receive input information for calibration between an image and a heterogeneous sensor, a matrix calculation unit configured to calculate a correlation for projection of information of the heterogeneous sensor, a projection unit configured to project the information of the heterogeneous sensor onto an image domain using the correlation, and a two-dimensional (2D) heterogeneous sensor fusion unit configured to perform stick calibration modeling and design and apply a stick calibration filter.

A device may receive sensor data and video data associated with a vehicle, and may process the sensor data, with a rule-based detector model, to determine whether a probability of a vehicle accident satisfies a first threshold. The device may preprocess acceleration data of the sensor data to generate calibrated acceleration data, and may process the calibrated acceleration data, with an anomaly detector model, to determine whether the calibrated acceleration data includes anomalies. The device may filter the sensor data to generate filtered sensor data, and may process the filtered sensor data and anomaly data, with a decision model, to determine whether the probability of the vehicle accident satisfies a second threshold. The device may process the filtered sensor data, the anomaly data, and the video data, with a machine learning model, to determine whether the vehicle accident has occurred, and may perform one or more actions.

A method for performing a computer-aided diagnosis (CAD) includes: acquiring a medical image set; generating a three-dimensional (3D) tumor distance map corresponding to the medical image set, each voxel of the tumor distance map representing a distance from the voxel to a nearest boundary of a primary tumor present in the medical image set; and performing neural-network processing of the medical image set to generate a predicted probability map to predict presence and locations of oncology significant lymph nodes (OSLNs) in the medical image set, wherein voxels in the medical image set are stratified and processed according to the tumor distance map.

Methods of determining the depth of a scene and associated systems are disclosed herein. In some embodiments, a method can include augmenting depth data of a scene captured with a depth sensor with depth data from one or more images of the scene. For example, the method can include capturing image data of the scene with a plurality of cameras. The method can further include generating a point cloud representative of the scene based on the depth data from the depth sensor and identifying a missing region of the point cloud, such as a region occluded from the view of the depth sensor. The method can then include generating depth data for the missing region based on the image data. Finally, the depth data for the missing region can be merged with the depth data from the depth sensor to generate a merged point cloud representative of the scene.