Wu, Fangyu, Yan, Shiyang, Smith, JS 
ORCID: 0000-0002-0212-2365 and Zhang, Bailing
(2019)
Vehicle Re-identification in Still Images: Application of Semi-supervised Learning and Re-ranking.
Signal Processing-Image Communication, 76.
pp. 261-271.
|
Text
Vehcile-Re-ID.pdf - Author Accepted Manuscript Download (17MB) | Preview |
Abstract
Vehicle re-identification (re-ID), namely, finding exactly the same vehicle from a large number of vehicle images, remains a great challenge in computer vision. Most existing vehicle re-ID approaches follow a fully supervised learning methodology, in which sufficient labeled training data is required. However, this limits their scalability to realistic applications, due to the high cost of data labeling. In this paper, we adopted a Generative Adversarial Network (GAN) to generate unlabeled samples and enlarge the training set. A semi supervised learning scheme with the Convolutional Neural Networks (CNN) was proposed accordingly, which assigns a uniform label distribution to the unlabeled images to regularize the supervised model and improve the performance of the vehicle re-ID system. Besides, an improved re-ranking method based on the Jaccard distance and k-reciprocal nearest neighbors is proposed to optimize the initial rank list. Extensive experiments over the benchmark datasets VeR1-776, VehicleID and VehicleReID have demonstrated that the proposed method outperforms the state-of-the-art approaches for vehicle re-ID.
| Item Type: | Article |
|---|---|
| Uncontrolled Keywords: | Vehicle re-identification, Convolutional neural networks, Semi-supervised learning, Re-ranking |
| Depositing User: | Symplectic Admin |
| Date Deposited: | 01 May 2019 07:37 |
| Last Modified: | 19 Jan 2023 00:52 |
| DOI: | 10.1016/j.image.2019.04.021 |
| Related URLs: | |
| URI: | https://livrepository.liverpool.ac.uk/id/eprint/3039095 |
Dimensions
Dimensions
