Person re-ID while Crossing Different Cameras: Combination of Salient-Gaussian Weighted BossaNova and Fisher Vector Encodings
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2017, Vol 8, Issue 9
Abstract
Person re-identification (re-ID) is a challenging task in the camera surveillance field, since it addresses the problem of re-identifying people across multiple non-overlapping cameras. Most of existing approaches have been concentrated on: 1) achieving a robust and effective feature representation; and 2) enforcing discriminative metric learning to predict if two images represent the same identity. In this context, we present a new approach for person re-ID built upon multi-level descriptors. This is achieved by combining three complementary representations: salient-Gaussian Fisher Vector (SGFV) encoding method, salient-Gaussian BossaNova (SGBN) histogram encoding method and deep Convolutional Neural Network (CNN) features. The two first methods adapt the histogram encoding framework to the person re-ID task. This is achieved by integrating the pedestrian saliency map and the spatial location information, in the histogram encoding process. On one hand, human saliency is reliable and distinctive in the person re-ID task, since it can model the uniqueness of the identity. On the other hand, localizing a person in the image can effectively discard noisy background information. Finally, one of the most advanced metric learning in person re-ID: the Cross-view Quadratic Discriminant Analysis (XQDA) is applied on the top of the resulting description. The proposed method yields promising person re-ID results on two challenging image-based person re-ID benchmarks: CUHK03 and Market-1501.
Authors and Affiliations
Mahmoud Mejdoub, Salma Ksibi, Chokri Ben Amar, Mohamed Koubaa
Keywords
Related Articles
- EP ID EP261222
- DOI 10.14569/IJACSA.2017.080953
- Views 70
- Downloads 0
Investigating Students’ Achievements in Computing Science Using Human Metric
This study investigates the role of personality traits, motivation for career choice and study habits in students’ academic achievements in the computing sciences. A quantitative research method was employed. Data was co...
LOQES: Model for Evaluation of Learning Object
Learning Object Technology is a diverse and contentious area, which is constantly evolving, and will inevitably play a major role in shaping the future of both teaching and learning. Learning Objects are small chun...
Data Mining for Engineering Schools
the supervision of the academic performance of engineering students is vital during an early stage of their curricula. Indeed, their grades in specific core/major courses as well as their cumulative General Point Average...
ADBT Frame Work as a Testing Technique: An Improvement in Comparison with Traditional Model Based Testing
Software testing is an embedded activity in all software development life cycle phases. Due to the difficulties and high costs of software testing, many testing techniques have been developed with the common goal of test...
Virtualization Implementation Model for Cost Effective & Efficient Data Centers
With advancement of technology things are becoming simpler and easier for us. Automatic systems are being preferred over manual system. This unit talks about the basic definitions needed to understand the Project better...