Meta-BNS FOR Adversarial Data-Free Quantization
- Resource Type
- Conference
- Authors
- Fu, Siming; Wang, Hualiang; Cao, Yuchen; Hu, Haoji; Peng, Bo; Tan, Wenming; Ye, Tingqun
- Source
- 2022 IEEE International Conference on Image Processing (ICIP) Image Processing (ICIP), 2022 IEEE International Conference on. :4038-4042 Oct, 2022
- Subject
- Computing and Processing
Signal Processing and Analysis
Quantization (signal)
Image processing
Games
Generators
Data models
Convergence
Data-free Quantization
Meta-BNS
Adversarial Explore
- Language
- ISSN
- 2381-8549
Data-free quantization has recently been a promising method to perform quantization without access to the original data. However, the drawback of such approaches is the homogenization of synthetic data due to low efficiency for diverse data generation and the performance collapse of the generator. To alleviate the above issue, we propose a novel Meta-BNS for adversarial data-free quantization scheme which consists of Meta-BNS module and adversarial exploration module. Meta-BNS module automatically learns an enhancement coefficient matrix function for BN loss module to provide a suitable constrain on the generator. Adversarial exploration module leverages minimax game between the generator and quantized model via input gradient to encourage the generator to learn high-dimensional and complex real data distribution. The experimental results show that our method achieves state-of-the-art performance for various settings on data-free quantization.