Class-Incremental Learning for Action Recognition in Videos
It addresses the problem of forgetting old classes when learning new ones in video recognition, which is incremental as it adapts continual learning techniques to a less-explored video domain.
The paper tackles catastrophic forgetting in class-incremental learning for video action recognition by introducing time-channel importance maps and a regularization scheme, achieving improved accuracy on benchmarks like UCF101, HMDB51, and Something-Something V2 compared to existing methods designed for images.
We tackle catastrophic forgetting problem in the context of class-incremental learning for video recognition, which has not been explored actively despite the popularity of continual learning. Our framework addresses this challenging task by introducing time-channel importance maps and exploiting the importance maps for learning the representations of incoming examples via knowledge distillation. We also incorporate a regularization scheme in our objective function, which encourages individual features obtained from different time steps in a video to be uncorrelated and eventually improves accuracy by alleviating catastrophic forgetting. We evaluate the proposed approach on brand-new splits of class-incremental action recognition benchmarks constructed upon the UCF101, HMDB51, and Something-Something V2 datasets, and demonstrate the effectiveness of our algorithm in comparison to the existing continual learning methods that are originally designed for image data.