Visualizing Deep Networks by Optimizing with Integrated Gradients
This work addresses the need for reliable interpretability in computer vision, particularly for users relying on heatmaps to understand deep learning models, though it is incremental as it builds on existing heatmap optimization techniques.
The paper tackled the problem of generating faithful heatmaps for visualizing deep networks by proposing I-GOS, which optimizes heatmaps using integrated gradients to reduce classification scores on masked images, resulting in heatmaps that are more correlated with network decisions compared to state-of-the-art methods.
Understanding and interpreting the decisions made by deep learning models is valuable in many domains. In computer vision, computing heatmaps from a deep network is a popular approach for visualizing and understanding deep networks. However, heatmaps that do not correlate with the network may mislead human, hence the performance of heatmaps in providing a faithful explanation to the underlying deep network is crucial. In this paper, we propose I-GOS, which optimizes for a heatmap so that the classification scores on the masked image would maximally decrease. The main novelty of the approach is to compute descent directions based on the integrated gradients instead of the normal gradient, which avoids local optima and speeds up convergence. Compared with previous approaches, our method can flexibly compute heatmaps at any resolution for different user needs. Extensive experiments on several benchmark datasets show that the heatmaps produced by our approach are more correlated with the decision of the underlying deep network, in comparison with other state-of-the-art approaches.