Poly-CAM: High resolution class activation map for convolutional neural networks
This work addresses the need for more accurate explainable AI tools in deep learning, particularly for image analysis, though it appears incremental by building on existing CAM techniques.
The paper tackles the problem of inaccurate localization in saliency maps for convolutional neural networks by proposing Poly-CAM, which combines information from earlier and later layers to produce high-resolution class activation maps. The result is competitive performance in faithfulness metrics and improved precision in localizing class-specific features compared to previous methods.
The need for Explainable AI is increasing with the development of deep learning. The saliency maps derived from convolutional neural networks generally fail in localizing with accuracy the image features justifying the network prediction. This is because those maps are either low-resolution as for CAM [Zhou et al., 2016], or smooth as for perturbation-based methods [Zeiler and Fergus, 2014], or do correspond to a large number of widespread peaky spots as for gradient-based approaches [Sundararajan et al., 2017, Smilkov et al., 2017]. In contrast, our work proposes to combine the information from earlier network layers with the one from later layers to produce a high resolution Class Activation Map that is competitive with the previous art in term of insertion-deletion faithfulness metrics, while outperforming it in term of precision of class-specific features localization.