Rethinking Class Activation Maps for Segmentation: Revealing Semantic Information in Shallow Layers by Reducing Noise
This work addresses a bottleneck in weakly supervised learning for segmentation by enhancing semantic information extraction, though it is incremental as it builds on existing CAM-related methods.
The paper tackled the problem of low-resolution class activation maps in weakly supervised semantic segmentation by proposing a gradient-based denoising method to filter noise in shallow feature maps, resulting in improved performance as demonstrated through extensive experiments.
Class activation maps are widely used for explaining deep neural networks. Due to its ability to highlight regions of interest, it has evolved in recent years as a key step in weakly supervised learning. A major limitation to the performance of the class activation maps is the small spatial resolution of the feature maps in the last layer of the convolutional neural network. Therefore, we expect to generate high-resolution feature maps that result in high-quality semantic information. In this paper, we rethink the properties of semantic information in shallow feature maps. We find that the shallow feature maps still have fine-grained non-discriminative features while mixing considerable non-target noise. Furthermore, we propose a simple gradient-based denoising method to filter the noise by truncating the positive gradient. Our proposed scheme can be easily deployed in other CAM-related methods, facilitating these methods to obtain higher-quality class activation maps. We evaluate the proposed approach through a weakly-supervised semantic segmentation task, and a large number of experiments demonstrate the effectiveness of our approach.