Reducing Information Bottleneck for Weakly Supervised Semantic Segmentation

TL;DR

This paper addresses the information bottleneck in weakly supervised semantic segmentation by removing the last activation function and introducing a new pooling method, significantly improving localization accuracy.

Contribution

It proposes a novel approach to reduce the information bottleneck in neural networks for semantic segmentation, leading to state-of-the-art results.

Findings

Improved localization maps on PASCAL VOC 2012

Enhanced performance on MS COCO 2014

Achieved new state-of-the-art results

Abstract

Weakly supervised semantic segmentation produces pixel-level localization from class labels; however, a classifier trained on such labels is likely to focus on a small discriminative region of the target object. We interpret this phenomenon using the information bottleneck principle: the final layer of a deep neural network, activated by the sigmoid or softmax activation functions, causes an information bottleneck, and as a result, only a subset of the task-relevant information is passed on to the output. We first support this argument through a simulated toy experiment and then propose a method to reduce the information bottleneck by removing the last activation function. In addition, we introduce a new pooling method that further encourages the transmission of information from non-discriminative regions to the classification. Our experimental evaluations demonstrate that this simple modification significantly improves the quality of localization maps on both the PASCAL VOC 2012 and MS COCO 2014 datasets, exhibiting a new state-of-the-art performance for weakly supervised semantic segmentation. The code is available at: https://github.com/jbeomlee93/RIB.