Knowledge Commons of Institute of Automation,CAS
Motion Complementary Network for Efficient Action Recognition | |
Cheng,Ke1,2; Zhang,Yifan1,2; Li,Chenghua1,2; Cheng,Jian1,2,3; Lu,Hanqing1,2 | |
2021-01 | |
会议名称 | International Conference on Pattern Recognition (ICPR) |
会议日期 | January 2021 |
会议地点 | 线上 |
摘要 | Both two-stream ConvNet and 3D ConvNet are widely used in action recognition. However, both methods are not efficient for deployment: calculating optical flow is very slow, while 3D convolution is computationally expensive. Our key insight is that the motion information from optical flow maps is complementary to the motion information from 3D ConvNet. Instead of simply combining these two methods, we propose two novel techniques to enhance the performance with less computational cost: fixed-motion-accumulation and balanced-motion-policy. With these two techniques, we propose a novel framework called Efficient Motion Complementary Network(EMC-Net) that enjoys both high efficiency and high performance. We conduct extensive experiments on Kinetics, UCF101, and Jester datasets. We achieve notably higher performance while consuming 4.7× less computation than I3D, 11.6× less computation than ECO, 17.8× less computation than R(2+1)D. On Kinetics dataset, we achieve 2.6% better performance than the recent proposed TSM with 1.4× fewer FLOPs and 10ms faster on K80 GPU. |
语种 | 英语 |
七大方向——子方向分类 | 图像视频处理与分析 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/45075 |
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 | Zhang,Yifan |
作者单位 | 1.NLPR & AIRIA, Institute of Automation, Chinese Academy of Sciences 2.School of Artificial Intelligence, University of Chinese Academy of Sciences 3.Research Center for Brain-inspired Intelligence, CASIA |
第一作者单位 | 模式识别国家重点实验室 |
通讯作者单位 | 模式识别国家重点实验室 |
推荐引用方式 GB/T 7714 | Cheng,Ke,Zhang,Yifan,Li,Chenghua,et al. Motion Complementary Network for Efficient Action Recognition[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 | ||
Motion Complementary(1128KB) | 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