In this paper we introduce a new computervision framework for the analysis and interpretation of the cephalo-ocular behavior of drivers. We start by detecting the most important facial features, namely the nose tip a...
详细信息
the proceedings contain 50 papers. the topics discussed include: probabilistic framework for feature-point matching;a PCA-based binning approach for matching to large SIFT database;matching maximally stable extremal r...
ISBN:
(纸本)9780769540405
the proceedings contain 50 papers. the topics discussed include: probabilistic framework for feature-point matching;a PCA-based binning approach for matching to large SIFT database;matching maximally stable extremal regions using edge information and the chamfer distance function;quasi-random scale space approach to robust keypoint extraction in high-noise environments;Bayesian identity clustering;automated classification of operational SAR sea ice images;efficient augmentation of the EKF structure from motion with frame-to-frame features;real-time virtual viewpoint generation on the GPU for scene navigation;a simple but effective approach to video copy detection;texture classification using compressed sensing;the effect of colour space on image sharpening algorithms;and automated filter parameter selection using measures of noiseness.
Scene classification is a hot topic in patternrecognition and computervision area. In this paper, based on the past research on vision neuroscience, we proposed a new biologically inspired feature method for scene i...
详细信息
the bag-of-visual-words (BOVW) approaches are widely used in human action recognition. Usually, large vocabulary size of the BOVW is more discriminative for inter-class action classification while small one is more ro...
详细信息
ISBN:
(纸本)9783642122965
the bag-of-visual-words (BOVW) approaches are widely used in human action recognition. Usually, large vocabulary size of the BOVW is more discriminative for inter-class action classification while small one is more robust to noise and thus tolerant to the intra-class invariance. In this pape, we propose a pyramid vocabulary tree to model local spatio-temporal features, which can characterize the inter-class difference and also allow intra-class variance. Moreover, since BOVW is geometrically unconstrained, we further consider the spatio-temporal information of local features and propose a sparse spatio-temporal pyramid matching kernel (termed as SST-PMK) to compute the similarity measures between video sequences. SST-PMK satisfies the Mercer's condition and therefore is readily integrated into SVM to perform action recognition. Experimental results on the Weizmann datasets show that boththe pyramid vocabulary tree and the SST-PMK lead to a significant improvement in human action recognition.
Behavior recognition is an attractive direction in the computervision domain. In this paper, we propose a novel behavior recognition method based on prototype learning using metric learning. Prototype learning algori...
详细信息
Several studies of psychophysics have shown that the eyes or the mouth seem to be an important cue in human face perception, and the nose plays an insignificant role, this means that there exists a distinctive informa...
详细信息
ISBN:
(纸本)9783642122965
Several studies of psychophysics have shown that the eyes or the mouth seem to be an important cue in human face perception, and the nose plays an insignificant role, this means that there exists a distinctive information distribution of faces. this paper presents a novel approach for face recognition by combining the Local Binary patterns (LBP) based face descriptor and the distinctive information of faces. First, we give a quantitative estimation of the density for each pixel in fronted face image by combining Parzen-window approach and Scale Invariant Feature Transform (SIFT) detector, which is taken as the measure of the distinctive information of the faces. Second, we integral the density function in the sub-window region of face to gain the weights set which is used in the LBP based face descriptor to produce weighted Chi square statistics. As an elementary application of the estimation of distinctive information of face, the proposed method is tested on the FERET FA/FB image sets and yields a recognition rate of 98.2% contrast to the 97.3% which is produced by the method adopted by Ahonen.
this paper presents a new action recognition approach based on local spatio-temporal features. the main contributions of our approach are twofold. First, a new local spatio-temporal feature is proposed to represent th...
详细信息
ISBN:
(纸本)9783642123061
this paper presents a new action recognition approach based on local spatio-temporal features. the main contributions of our approach are twofold. First, a new local spatio-temporal feature is proposed to represent the cuboids detected in video sequences. Specifically, the descriptor utilizes the covariance matrix to capture the self-correlation information of the low-level features within each cuboid. Since covariance matrices do not lie on Euclidean space, the Log-Euclidean Riemannian metric is used for distance measure between covariance matrices. Second, the Earth Mover's Distance (EMD) is used for matching any pair of video sequences. In contrast to the widely used Euclidean distance, EMD achieves more robust performances in matching histograms/distributions with different sizes. Experimental results on two datasets demonstrate the effectiveness of the proposed approach.
In this report we introduce a novel approach for determining correspondence in a sequence of images. We formulate a probabilistic framework that relates a feature's appearance and its position under relaxed statis...
详细信息
this paper proposes a local motion-based approach for recognizing group activities in soccer videos. Given the SIFT keypoint matches on two successive frames, we propose a simple but effective method to group these ke...
详细信息
ISBN:
(纸本)9783642123061
this paper proposes a local motion-based approach for recognizing group activities in soccer videos. Given the SIFT keypoint matches on two successive frames, we propose a simple but effective method to group these keypoints into the background point set and the foreground point set. the former one is used to estimate camera motion and the latter one is applied to represent group actions. After camera motion compensation, we apply a local motion descriptor to characterize relative motion between corresponding keypoints on two consecutive frames. the novel descriptor is effective in representing group activities since it focuses on local motion of individuals and excludes noise such as background motion caused by inaccurate compensation. Experimental results show that our approach achieves high recognition rates in soccer videos and is robust to inaccurate compensation results.
this paper proposes an automatic algorithm to segment multiple objects from the initial frame of the multi-view sequence. Firstly, saliency based visual attention model is adopted in the key view of image to detect th...
详细信息
ISBN:
(纸本)9780889868236
this paper proposes an automatic algorithm to segment multiple objects from the initial frame of the multi-view sequence. Firstly, saliency based visual attention model is adopted in the key view of image to detect the location of objects-of-interest (OOIs), followed by the OOIs extraction based on the saliency analysis. then, the initial mask of the key view is spatially projected by the disparity value to form the ones of the other views. Finally, multiple objects segmentation is decomposed into several sub-segmentation problems and solved by minimizing our proposed energy function using bi-label graph cut to achieve more accurate objects representation. Comparing with a related algorithm on our captured multi-view sequence, experimental results demonstrate the advantage of the proposed algorithm in terms of computational efficiency and segmentation performance.
暂无评论