Kaifang Li, Guancheng Hui, Ruhan Wang, Miaohui Zhang. Person Re-Identification Based on Generative Adversarial Network and Self-Calibrated Convolution[J]. Laser & Optoelectronics Progress, 2022, 59(10): 1015007
- Laser & Optoelectronics Progress
- Vol. 59, Issue 10, 1015007 (2022)
Fig. 1. GAN diagram
Fig. 2. Cyclic network working principle. (a) Generator and discriminator; (b) forward cycle-consistency loss; (c) backward cycle-consistency loss
Fig. 3. Examples generated by CycleGAN and CVQGAN in Market-1501
Fig. 4. Flow chart of CVQGAN generator
Fig. 5. CVQStyle image generation network
Fig. 6. Self-calibrated convolution module network structure
Fig. 7. Overall network framework
Fig. 8. Image examples generated by DCGAN, CycleGAN, and CVQGAN.
Fig. 9. Visualization results of Market-1501 dataset. (a) (c) Reference model; (b) (d) proposed model
|
Table 1. Generated image quality comparison
| ||||||||||||||||||||||||||||||||||
Table 2. Experimental results of different models
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Table 3. Performance comparison of proposed algorithm with mainstream algorithms on Market-1501 and DukeMTMC-reID datasets
Set citation alerts for the article
Please enter your email address
© Copyright 2018-2021 | Chinese Laser Press. All Rights Reserved 沪ICP备15018463号-20