DecomCAM: Advancing Beyond Saliency Maps through Decomposition and Integration

TL;DR

DecomCAM introduces a novel decomposition and integration approach using singular value decomposition to produce detailed, interpretable saliency maps for complex deep networks, improving localization accuracy and interpretability.

Contribution

It presents DecomCAM, a new method that decomposes activation maps into orthogonal sub-saliency maps for better interpretability of deep models.

Findings

Outperforms existing methods in localization accuracy.

Balances interpretability with computational efficiency.

Reveals object components aligned with model reasoning.

Abstract

Interpreting complex deep networks, notably pre-trained vision-language models (VLMs), is a formidable challenge. Current Class Activation Map (CAM) methods highlight regions revealing the model's decision-making basis but lack clear saliency maps and detailed interpretability. To bridge this gap, we propose DecomCAM, a novel decomposition-and-integration method that distills shared patterns from channel activation maps. Utilizing singular value decomposition, DecomCAM decomposes class-discriminative activation maps into orthogonal sub-saliency maps (OSSMs), which are then integrated together based on their contribution to the target concept. Extensive experiments on six benchmarks reveal that DecomCAM not only excels in locating accuracy but also achieves an optimizing balance between interpretability and computational efficiency. Further analysis unveils that OSSMs correlate with discernible object components, facilitating a granular understanding of the model's reasoning. This positions DecomCAM as a potential tool for fine-grained interpretation of advanced deep learning models. The code is avaible at https://github.com/CapricornGuang/DecomCAM.