arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2203.13611cs.CV
收藏

类增量学习在视频动作识别中的应用

Class-Incremental Learning for Action Recognition in Videos

Seoul National University(首尔大学)

Jaeyoo Park, Minsoo Kang, Bohyung Han

摘要:

我们解决灾难性遗忘问题的背景下,类增量学习的视频识别,这还没有被积极探索,尽管流行的持续学习。我们的框架解决了这一具有挑战性的任务,通过引入时间通道的重要性地图,并利用重要性地图通过知识蒸馏学习传入的例子的表示。我们还在目标函数中加入了正则化方案,该方案鼓励从视频中不同时间步获得的各个特征不相关,并最终通过减轻灾难性遗忘来提高准确性。我们评估所提出的方法上构建的UCF 101,HMDB 51,和Something-Something V2数据集的类增量动作识别基准的全新的分裂,并证明我们的算法的有效性相比,现有的持续学习方法,最初是为图像数据设计的。

英文摘要:

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.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑