首页|Behavior detection and evaluation based on multi‑frame MobileNet
Behavior detection and evaluation based on multi‑frame MobileNet
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Springer Nature
Video-based behavior detection is an important research direction in computer vision,which has great application potential in intelligent video surveillance, sports behavior evaluation,gait recognition, and so on. However, due to the complexity of video content andbackground, video behavior detection and evaluation face many challenges and are stillin their early stages. This paper proposes a novel multi-frame MobileNet model, whichdescribes the internal differences of similar behaviors by introducing multiple continuousframes of behaviors to be detected, and realizes fine-grained behavior detection and evaluation.Firstly, using energy trend images (ETIs) of behaviors as features, multiple continuousframes of the target video are fed into the proposed network to explore the relationshipbetween adjacent frames. Then,in the weighted point-wise convolution stage, by adding afade-in factor to the timeline for providing different weights to each involved frame, whichmakes better use of the progressive relationship between behavior frames at different times.Finally, the effectiveness of the proposed method is verified by comparative experimentson multiple video data sets such as UCF101, HMDB51 and CASIA-B.
NETL
NSTL
Springer Nature
