Academic
Publications
Person Re-identification Using Spatial Covariance Regions of Human Body Parts

Person Re-identification Using Spatial Covariance Regions of Human Body Parts,10.1109/AVSS.2010.34,Slawomir Bak,E. Corvee,F. Brémond,M. Thonnat

Person Re-identification Using Spatial Covariance Regions of Human Body Parts   (Citations: 4)
BibTex | RIS | RefWorks Download
In many surveillance systems there is a requirement to determine whether a given person of interest has already been observed over a network of cameras. This is the person re-identification problem. The human appearance obtained in one camera is usually different from the ones obtained in another camera. In order to re-identify people the human signature should handle difference in illumination, pose and camera parameters. We propose a new appearance model based on spatial covariance regions extracted from human body parts. The new spatial pyramid scheme is applied to capture the correlation between human body parts in order to obtain a discriminative human signature. The human body parts are automatically detected using Histograms of Oriented Gradients (HOG). The method is evaluated using benchmark video sequences from i-LIDS Multiple-Camera Tracking Scenario data set. The re-identification performance is presented using the cumulative matching characteristic (CMC) curve. Finally, we show that the proposed approach outperforms state of the art methods.
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
    • ...(iii) Compared to most other state-of-the-art approaches like [2, 3], this approach is applicable in real applications since it is integrated with a detection and tracking strategy...
    • ...Particularly, it does not rely on manual annotation of people like [2, 3] and builds models for re-identification online without an offline training step like [1, 5]...
    • ...In detail, following approaches are shown: Spatial covariance Regions (SCR) [2], Center Rectangular Ring Ratio-Occurrence Descriptor-Block based Ratio-Occurrence (CRRRO-BRO), Appearance Context, Color histogram and Shape and Appearance Context from [15]...

    Kai Junglinget al. Person re-identification in multi-camera networks

    • ...We compared our approach with methods which obtained the best performance on this dataset: SCR [2], HPE...
    • ...For each pedestrian we create human signature using N = 1 (for SCR [2]) or N = 10 (for M-SCR and MRCG) randomly selected images...
    • ...Figure 4. CMC curves obtained on i-LIDS datasets; (a) We compare our methods with the results of SCR [2] and M-SCR, HPE [3], SDALF [6] and Appearance context model with Group Context [21]; (b) and (c) We compare our methods with the results of SCR [2] and M-SCR...
    • ...Figure 4. CMC curves obtained on i-LIDS datasets; (a) We compare our methods with the results of SCR [2] and M-SCR, HPE [3], SDALF [6] and Appearance context model with Group Context [21]; (b) and (c) We compare our methods with the results of SCR [2] and M-SCR...
    • ...These methods are named single- shot approaches [2, 15, 20] and until now they were the most popular techniques...

    Slawomir Baket al. Multiple-shot human re-identification by Mean Riemannian Covariance Gr...

    • ...(iii) Compared to most other state-of-the-art approaches like [3, 4], this approach is applicable in real applications since it is integrated with a detection and tracking strategy...
    • ...Specifically, it does not rely on manual annotation of people like [3, 4] and builds models for re-identification online without an offline training step like [2, 5]...

    Kai Junglinget al. View-invariant person re-identification with an Implicit Shape Model

    • ...Recently, a covariance descriptor [9] has proved its effectiveness in recognition [1] and classification approaches [8]...
    • ...It has been shown that the performance of the covariance features is superior to other methods, as rotation and illumination changes are absorbed by the covariance matrix [1]...
    • ...In our approach a human appearance is represented by a set of covariance matrices extracted on different resolutions from detected body parts [1]...

    Sławomir Bąket al. Human Re-identification System on Highly Parallel GPU and CPU Architec...

Sort by: