Information-Bottleneck Approach to Salient Region Discovery
This work addresses the challenge of salient region discovery for image classification, but it appears incremental as it builds on existing attention methods with a Boolean mask approach.
The paper tackles the problem of learning image attention masks in a semi-supervised setting by proposing a method based on the Information Bottleneck principle, which minimizes mutual information between input and masked images while maximizing it between masked images and labels, and demonstrates successful feature attention on synthetic datasets like MNIST, CIFAR10, and SVHN.
We propose a new method for learning image attention masks in a semi-supervised setting based on the Information Bottleneck principle. Provided with a set of labeled images, the mask generation model is minimizing mutual information between the input and the masked image while maximizing the mutual information between the same masked image and the image label. In contrast with other approaches, our attention model produces a Boolean rather than a continuous mask, entirely concealing the information in masked-out pixels. Using a set of synthetic datasets based on MNIST and CIFAR10 and the SVHN datasets, we demonstrate that our method can successfully attend to features known to define the image class.