Knowledge Commons of Institute of Automation,CAS
Generative Zero-shot Network Quantization | |
Xiangyu, He1,2![]() ![]() ![]() ![]() ![]() ![]() | |
2021-06 | |
会议名称 | IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR) 2021 Workshops |
会议日期 | 2021-6 |
会议地点 | Virtual Event |
摘要 | Convolutional neural networks are able to learn realistic image priors from numerous training samples in low-level image generation and restoration [66]. We show that, for high-level image recognition tasks, we can further reconstruct “realistic” images of each category by leveraging intrinsic Batch Normalization (BN) statistics without any training data. Inspired by the popular VAE/GAN methods, we regard the zero-shot optimization process of synthetic images as generative modeling to match the distribution of BN statistics. The generated images serve as a calibration set for the following zero-shot network quantizations. Our method meets the needs for quantizing models based on sensitive information, e.g., due to privacy concerns, no data is available. Extensive experiments on benchmark datasets show that, with the help of generated data, our approach consistently outperforms existing data-free quantization methods. |
语种 | 英语 |
七大方向——子方向分类 | 其他 |
国重实验室规划方向分类 | 其他 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/48941 |
专题 | 复杂系统认知与决策实验室_高效智能计算与学习 |
通讯作者 | Jian, Cheng |
作者单位 | 1.Institute of Automation, Chinese Academy of Sciences 2.University of Chinese Academy of Scienses |
第一作者单位 | 中国科学院自动化研究所 |
通讯作者单位 | 中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Xiangyu, He,Jiahao, Lu,Weixiang, Xu,et al. Generative Zero-shot Network Quantization[C],2021. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
Generative_Zero-Shot(947KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论