Towards Weakly Supervised End-to-end Learning for Long-video Action Recognition
This work addresses the high annotation cost for long-video action recognition, enabling scalable model training, though it is incremental as it builds on existing weakly supervised methods.
The paper tackles the problem of expensive action interval annotations for training end-to-end action recognition models on long videos by proposing AdaptFocus, a weakly supervised framework that uses only video-level labels to adaptively focus on informative clips, achieving state-of-the-art results on three datasets and advancing downstream tasks.
Developing end-to-end action recognition models on long videos is fundamental and crucial for long-video action understanding. Due to the unaffordable cost of end-to-end training on the whole long videos, existing works generally train models on short clips trimmed from long videos. However, this ``trimming-then-training'' practice requires action interval annotations for clip-level supervision, i.e., knowing which actions are trimmed into the clips. Unfortunately, collecting such annotations is very expensive and prevents model training at scale. To this end, this work aims to build a weakly supervised end-to-end framework for training recognition models on long videos, with only video-level action category labels. Without knowing the precise temporal locations of actions in long videos, our proposed weakly supervised framework, namely AdaptFocus, estimates where and how likely the actions will occur to adaptively focus on informative action clips for end-to-end training. The effectiveness of the proposed AdaptFocus framework is demonstrated on three long-video datasets. Furthermore, for downstream long-video tasks, our AdaptFocus framework provides a weakly supervised feature extraction pipeline for extracting more robust long-video features, such that the state-of-the-art methods on downstream tasks are significantly advanced. We will release the code and models.