Top-down Neural Attention by Excitation Backprop
This work addresses the need for interpretable and discriminative attention mechanisms in CNNs for computer vision tasks, offering a novel method that is incremental but improves upon existing approaches.
The paper tackles the problem of generating task-specific attention maps for Convolutional Neural Networks by proposing Excitation Backprop, a new backpropagation scheme inspired by top-down human visual attention, and demonstrates its accuracy in weakly supervised localization tasks on datasets like MS COCO, PASCAL VOC07, and ImageNet, with promising performance in phrase localization on the Flickr30k Entities dataset.
We aim to model the top-down attention of a Convolutional Neural Network (CNN) classifier for generating task-specific attention maps. Inspired by a top-down human visual attention model, we propose a new backpropagation scheme, called Excitation Backprop, to pass along top-down signals downwards in the network hierarchy via a probabilistic Winner-Take-All process. Furthermore, we introduce the concept of contrastive attention to make the top-down attention maps more discriminative. In experiments, we demonstrate the accuracy and generalizability of our method in weakly supervised localization tasks on the MS COCO, PASCAL VOC07 and ImageNet datasets. The usefulness of our method is further validated in the text-to-region association task. On the Flickr30k Entities dataset, we achieve promising performance in phrase localization by leveraging the top-down attention of a CNN model that has been trained on weakly labeled web images.