CMSN: Continuous Multi-stage Network and Variable Margin Cosine Loss for Temporal Action Proposal Generation
This work addresses the problem of temporal action proposal generation for video analysis, but it is incremental as it builds on existing network architectures and datasets.
The paper tackles the challenge of accurately locating action boundaries in untrimmed videos by proposing a Continuous Multi-stage Network (CMSN) and Variable Margin Cosine Loss (VMCL), achieving better performance than state-of-the-art methods on THUMOS14.
Accurately locating the start and end time of an action in untrimmed videos is a challenging task. One of the important reasons is the boundary of action is not highly distinguishable, and the features around the boundary are difficult to discriminate. To address this problem, we propose a novel framework for temporal action proposal generation, namely Continuous Multi-stage Network (CMSN), which divides a video that contains a complete action instance into six stages, namely Backgroud, Ready, Start, Confirm, End, Follow. To distinguish between Ready and Start, End and Follow more accurately, we propose a novel loss function, Variable Margin Cosine Loss (VMCL), which allows for different margins between different categories. Our experiments on THUMOS14 show that the proposed method for temporal proposal generation performs better than the state-of-the-art methods using the same network architecture and training dataset.