Abstract: Provided are a method and device for determining a lean angle of a body and a pose estimation method and device. The method for determining a lean angle of a body of the present invention includes: a head-position obtaining step for obtaining a position of a head; a search region determination step for determining a plurality of search region spaced with an angle around the head; an energy function calculating step for calculating a value of an energy function for the search region; and a lean angle determining step for determining the lean angle of a search region with a largest or smallest value of the energy function as the lean angle of the body. The pose estimation method of the present invention includes a body lean-angle obtaining step, for obtaining a lean angle of a body; and a pose estimation step, for performing a pose estimation based on the lean angle of the body.

Abstract: A method and apparatus that classify an image. The method extracts a feature vector from the image, wherein the feature vector includes a plurality of first features. The extracting of each of the first features includes: acquiring a difference between sums or mean values of pixels of the plurality of first areas in the corresponding combination to obtain a first difference vector in the direction of the first axis, and obtaining a second difference vector in the direction of the second axis. A first projection difference vector is acquired with a second projection difference vector. A sum of magnitudes of the first projection difference vector and the second projection difference vector as the first feature is obtained; and the image according to the extracted feature vector is classified.

Abstract: The present invention discloses a method and an apparatus for extracting a prosodic feature of a speech signal, the method including: dividing the speech signal into speech frames; transforming the speech frames from time domain to frequency domain; and extracting respective prosodic features for different frequency ranges. According to the above technical solution of the present invention, it is possible to effectively extract the prosodic feature which can combine with a traditional acoustics feature without any obstacle.

Abstract: A method and system for carrying out reliability classification for motion vectors in a video is proposed in the application. The method comprises: partitioning and searching step for partitioning a specified video frame of an input video, and searching motion vectors for a specified block of the specified video frame so as to generate a block matching error for the specified block; texture feature extracting step for extracting a texture feature of the specified block; and classifying-by-block step for carrying out reliability classification for the motion vectors for the specified block in accordance with the block matching error and the texture feature of the specified block.

Abstract: The present invention provides a method for detecting a specific object in an image to be detected, including: a feature extraction step for extracting an image feature of the image to be detected; and a detection step for detecting detection windows with various sizes of the image to be detected according to the extracted image feature by using classifiers with various sizes corresponding to at least a part of the detection windows with various sizes, so as to determine the presence and location of a specific object in the image to be detected. The invention further provides an object detection device and a system including the device. The method, device and system for detecting a specific object in an image to be detected can improve the precision and increase the speed of the object detection.

Abstract: A motion capture apparatus includes: generating means; predicting means; projecting means; evaluating means; and estimating means. The generating means generates a kinematics model having joints connected to one another. The predicting means hierarchically calculates predicted positions at a current time of the joints of the objective person, and predicts three-dimensional positions and postures of the joints and the limbs. The projecting means projects the three-dimensional positions and postures of the joints and the limbs of the objective person on a two-dimensional image plane, the three-dimensional positions and the postures being predicted by predicting means. The evaluating means evaluates reliability about projection positions projected by the projecting means based on the observation image of the objective person.

Abstract: The present invention provides a foreground action estimating apparatus and a foreground action estimating method, wherein the foreground action estimating apparatus includes: a training image inputting means for inputting a foreground image, a background image and an image having the foreground and background images as training images; a basis matrix calculating means for calculating a foreground basis matrix and a background basis matrix by respectively extracting a foreground feature and a background feature from the foreground image and the background image, respectively, and combining the foreground basis matrix and the background basis matrix to obtain a combined basis matrix; a feature suppressing means for calculating the feature coefficients of the training images in accordance with the combined basis matrix obtained by the basis matrix calculating means so as to obtain image features of the background-feature-suppressed training images; and a foreground action information acquiring means for estimat

Abstract: An object recognizing apparatus and method are provided. The apparatus may include: a viewing direction estimating device configured for respectively estimating a first viewing direction of a first object captured by a first camera and a second viewing direction of a second object captured by a second camera; a feature extracting device configured for extracting one or more features respectively from an image containing the first object captured by the first camera and an image containing the second object captured by the second camera; and an object matching device configured for allocating a weight for each of the one or more features according to the first viewing direction and the second viewing direction, and calculating a similarity between the first object and the second object based on the one or more weighted features, to determine whether the first object and the second object are the same object.

Abstract: A method of monitoring an object in images captured by N camera apparatuses including: for an ith camera apparatus among the N camera apparatuses, obtaining respective first matching similarities of a specific object in an image captured by the ith camera apparatus with respect to one or more objects in an image captured by a jth camera apparatus respectively according to a pre-constructed feature conversion model between the camera apparatuses; and determining an object matching with the specific object in the image captured by the jth camera apparatus based on the respective first matching similarities to thereby monitor the specific object. There is further disclosed a method of performing an interactive operation of a related monitored object by using the foregoing monitoring method.

Abstract: The disclosure provides abnormal behavior detecting apparatus and method. The apparatus may include: an extracting device configured to extract, from a video segment to be detected, an image block sequence containing a plurality of image blocks corresponding to a moving range of an object in each image frame in the video segment; a feature calculating device configured to calculate motion vector features of the image block sequence; and an abnormal behavior detecting device comprising two or more stages of classifiers that are connected in series. The classifiers are configured to receive the image block sequence and the motion vector features stage by stage and detect the abnormal behavior of the object. If a previous stage of classifier determines that the to image block sequence contains an abnormal behavior, a next stage of classifier further receives and detects the image block sequence, until last stage of classifier.

Abstract: The present invention provides a moving image analyzing apparatus, method and system. The moving image analyzing apparatus comprises a moving image reading means for reading a moving image, a region-of-interest extracting means for extracting a region-of-interest from each frame in the moving image, an object feature extracting means for extracting an object feature in the region-of-interest or a region adjacent to the region-of-interest, and a shot change detecting means for detecting a shot change on the basis of the color feature of image, the object feature of the region-of-interest, and the differences of the motion information among the images of each frame. By estimating the reliability of the motion information within an image, the present invention can eliminate the shot change which is incorrectly detected according to the color distribution feature and the dimensional feature of the region-of-interest, thereby improving the detection accuracy of shot change.

Abstract: An object detecting apparatus and method is disclosed. An object detecting apparatus comprises: a detection classifier, configured to detect an object in an input image to obtain one or more candidate objects; a verifying classifier, configured to verify each candidate object by using verifying features from an image block corresponding to the each candidate object; and an on-line learning device, configured to train and optimize the detection classifier by using image blocks corresponding to the candidate objects as on-line samples, based on verifying results of the candidate objects obtained by the verifying classifier.

Abstract: The present invention discloses a system and method for segmenting foreground and background in a video, wherein the system comprises: a video shooting module for shooting the video; a data reading module for reading each frame of the video; a primary segmentation module for establishing a plurality of Gaussian models in a first color space for each pixel of each frame, and performing a matching processing between each pixel of the current frame and the plurality of Gaussian models corresponding to the pixel, and primarily segmenting the pixels as foreground and background according to the result of the matching processing; and a segmentation rejudging module for performing a rejudging processing to the primarily segmented foreground and background in a second color space, so as to obtain the finally determined foreground and background. The present invention improves the effect of foreground segmentation by using the combination of the color spaces and introducing the relationship between pixels.

Abstract: A method and apparatus for evaluating a human pose recognition technology, where this method comprises: a loading step of loading a pose recognition module with the human pose recognition technology to be evaluated; a pose recognizing step of recognizing a human pose for each test image in a test image set using the pose recognition module to obtain pose data for each test image; a pose classifying step of classifying the pose data for each test image in terms of a predetermined pose category set to obtain a pose category value for each test image; a pose classification evaluating step of calculating a pose classification accuracy rate by comparing the pose category value for each test image with a pose category true value; and a pose recognition comprehensive evaluating step of calculating a comprehensive evaluation score for the human pose recognition technology according to the pose classification accuracy rate.

Abstract: An audio event detection method and apparatus based on the long-term feature is provided. The audio event detection method comprises the step: dividing the input audio stream into a series of slices; extracting the short-term features and the long-term features for each slice; and obtaining the classification result of the input audio stream based on the short-term features and the long-term features.

Abstract: An object detection and classification method and apparatus, the method includes: inputting an image; performing window scan on the image to make, for each class, object existence decision on each window, and making, on a window of which the decision is positive, an object class decision to obtain an object classification confidence of the window; performing, for all classes, spatial neighborhood merging on all windows with positive result of object existence decision to obtain merged regions and object detection confidences thereof; judging, for each merged region, whether the object detection confidence of the merged region is higher than a predetermined threshold; calculating, for each class, if higher than the predetermined threshold, a merged object classification confidence for all the windows with positive result of object existence decision within the merged region; and determining a class with a highest merged object classification confidence as the class of the merged region.

Abstract: A method and a device for classifying an image are provided. The method includes: extracting a set of features as a feature vector, wherein the extracting includes: for each feature of the feature vector, determining a plurality of first areas arranged along a first axis and a plurality of second areas arranged along a second axis intersecting with the first axis; calculating the first differences between the sums of the pixels or the mean values of the plurality of first areas and the second differences between the sums of the pixels or the mean values of the plurality of second areas; and calculating the magnitude of gradient and the direction of gradient based on the first differences and the second differences, so as to form each feature; and according to the extracted feature vector, classifying the image.

Abstract: A device for detecting a specific object includes: a suspect object region detection unit configured to create a foreground mask of each frame of image in a sequence of images and perform an inter-frame differential process on the foreground masks to detect a suspect object region; a unit for modeling a region with high incidence of false positive configured to, if at least one suspect object region is detected, determine a suspect object region satisfying a predetermined condition as a region with high incidence of false positive and build a model of each determined region; a post-processing unit configured to match each suspect object region not determined as a region with high incidence of false positive against at least one corresponding model, and detect the specific object according to a sequence of mismatching suspect object regions; and determine absence of the specific object if no suspect object region is detected.

Abstract: The present invention provides a device and method for multiclass object detection, wherein the detection device includes: an input unit configured to input data to be detected; and a joint classifier within which a plurality of strong classifiers capable of processing multiclass object data are included, wherein each of the strong classifiers is acquired by adding a set of weak classifiers together, and each weak classifiers performs a weak classification for the data to be detected by using a feature. A list of shared features is included within the joint classifier, and each feature within the list is shared by one or more weak classifiers belonging to different strong classifiers respectively; and the weak classifiers, which use a same feature and belong to different strong classifiers respectively, have different parameter values from one another.

Abstract: The present invention discloses a three dimensional human pose recognition method and apparatus, where the three dimensional human pose recognition method includes steps of: a three dimensional pose initial recognition step of performing three dimensional pose recognition on an input image containing a human image to obtain image-based three dimensional human pose information; a sensor information acquisition step of acquiring, by a motion sensor, motion information of human articulation points; and a three dimensional pose correction step of correcting, with the motion information acquired by the sensor information acquisition step, the image-based three dimensional human pose information recognized by the three dimensional pose initial recognition step. According to the technical solution of the invention, it is possible to improve accuracy of the three dimensional human pose recognition efficiently.