Why Can't I Dance in the Mall? Learning to Mitigate Scene Bias in Action Recognition
Why Can't I Dance in the Mall? Learning to Mitigate Scene Bias in Action Recognition
复制标题
DOI：
--
复制
发表时间：
2019-12
期刊：
影响因子：
--
通讯作者：
Jinwoo Choi;Chen Gao;Joseph C.E. Messou;Jia-Bin Huang
中科院分区：
文献类型：
--
作者：
Jinwoo Choi;Chen Gao;Joseph C.E. Messou;Jia-Bin Huang
人类活动通常发生在特定场景的情况下，例如在篮球场上打篮球。因此，使用现有视频数据集训练模型，因此不可避免地捕获和利用这种偏见（而不是使用实际的歧视性提示）。学习的表示形式可能无法很好地推广到新的行动类别或不同的任务。在本文中，我们建议减轻视频表示学习的场景偏见。具体而言，我们通过1）场景类型的对抗性损失增加了行动分类的标准跨凝结损失，以及2）对人类演员被掩盖的视频的人掩模混乱损失。这两种损失鼓励学习表征，当没有证据时，无法预测场景类型和正确的行动。我们通过将预训练的模型转移到三个不同的任务，包括动作分类，时间定位和时空动作检测来验证方法的有效性。我们的结果显示出对基线模型的一致改进而无需进行辩护。
Human activities often occur in specific scene contexts, e.g., playing basketball on a basketball court. Training a model using existing video datasets thus inevitably captures and leverages such bias (instead of using the actual discriminative cues). The learned representation may not generalize well to new action classes or different tasks. In this paper, we propose to mitigate scene bias for video representation learning. Specifically, we augment the standard cross-entropy loss for action classification with 1) an adversarial loss for scene types and 2) a human mask confusion loss for videos where the human actors are masked out. These two losses encourage learning representations that are unable to predict the scene types and the correct actions when there is no evidence. We validate the effectiveness of our method by transferring our pre-trained model to three different tasks, including action classification, temporal localization, and spatio-temporal action detection. Our results show consistent improvement over the baseline model without debiasing.