论文标题
对象 - ABN：学会生成尖锐的注意图以进行行动识别
Object-ABN: Learning to Generate Sharp Attention Maps for Action Recognition
论文作者
论文摘要
在本文中，我们通过使用实例分割来生成更尖锐的注意图以进行动作识别，提出了注意力分支网络（ABN）的扩展。视觉解释的方法（例如Grad-CAM）通常会产生模糊的地图，这些图对人类的理解不是直观的，尤其是在识别视频中人们的行为时。我们提出的方法ABN通过引入新的面具丢失来解决此问题，该掩模损失使生成的注意图接近实例分段结果。此外，引入了PC丢失和多个注意图，以增强地图的清晰度并提高分类的性能。 UCF101和SSV2的实验结果表明，通过所提出的方法生成的地图在定性和定量上比原始ABN的图更清晰。
In this paper we propose an extension of the Attention Branch Network (ABN) by using instance segmentation for generating sharper attention maps for action recognition. Methods for visual explanation such as Grad-CAM usually generate blurry maps which are not intuitive for humans to understand, particularly in recognizing actions of people in videos. Our proposed method, Object-ABN, tackles this issue by introducing a new mask loss that makes the generated attention maps close to the instance segmentation result. Further the PC loss and multiple attention maps are introduced to enhance the sharpness of the maps and improve the performance of classification. Experimental results with UCF101 and SSv2 shows that the generated maps by the proposed method are much clearer qualitatively and quantitatively than those of the original ABN.
