Eigen-CAM: Class Activation Map using Principal Components

Abstract

Deep neural networks are ubiquitous due to the ease of developing models and their influence on other domains. At the heart of this progress is convolutional neural networks (CNNs) that are capable of learning representations or features given a set of data. Making sense of such complex models (i.e., millions of parameters and hundreds of layers) remains challenging for developers as well as the end-users. This is partially due to the lack of tools or interfaces capable of providing interpretability and transparency. A growing body of literature, for example, class activation map (CAM), focuses on making sense of what a model learns from the data or why it behaves poorly in a given task. This paper builds on previous ideas to cope with the increasing demand for interpretable, robust, and transparent models. Our approach provides a simpler and intuitive (or familiar) way of generating CAM. The proposed Eigen-CAM computes and visualizes the principle components of the learned features/representations from the convolutional layers. Empirical studies were performed to compare the Eigen-CAM with the state-of-the-art methods (such as Grad-CAM, Grad-CAM++, CNN-fixations) by evaluating on benchmark datasets such as weakly-supervised localization and localizing objects in the presence of adversarial noise. Eigen-CAM was found to be robust against classification errors made by fully connected layers in CNNs, does not rely on the backpropagation of gradients, class relevance score, maximum activation locations, or any other form of weighting features. In addition, it works with all CNN models without the need to modify layers or retrain models. Empirical results show up to 12% improvement over the best method among the methods compared on weakly supervised object localization.

arXiv comment7 pages, 4 figures

References in corpus (8)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Grad-CAM++: Improved Visual Explanations for Deep Convolutional Networks
• Fully Convolutional Networks for Semantic Segmentation
• Striving for Simplicity: The All Convolutional Net
• Learning Important Features Through Propagating Activation Differences
• Show and Tell: Lessons learned from the 2015 MSCOCO Image Captioning Challenge
• Fully Convolutional Networks for Semantic Segmentation
• Fast R-CNN

Cited by in corpus (16)

• Guided Hybrid Quantization for Object detection in Multimodal Remote Sensing Imagery via One-to-one Self-teaching
• Occlusion Sensitivity Analysis with Augmentation Subspace Perturbation in Deep Feature Space
• An Explainable Contrastive-based Dilated Convolutional Network with Transformer for Pediatric Pneumonia Detection
• An Open API Architecture to Discover the Trustworthy Explanation of Cloud AI Services
• RankNEAT: Outperforming Stochastic Gradient Search in Preference Learning Tasks
• Spatial Transformer Network YOLO Model for Agricultural Object Detection
• Learning Vision Transformer with Squeeze and Excitation for Facial Expression Recognition
• Towards explainable artificial intelligence (XAI) for early anticipation of traffic accidents
• Rethinking Person Re-Identification via Semantic-Based Pretraining
• PEPL: Precision-Enhanced Pseudo-Labeling for Fine-Grained Image Classification in Semi-Supervised Learning
• Towards Natural Machine Unlearning
• MedicalPatchNet: A Patch-Based Self-Explainable AI Architecture for Chest X-ray Classification
• Towards the Visualization of Aggregated Class Activation Maps to Analyse the Global Contribution of Class Features
• ComFace: Facial Representation Learning with Synthetic Data for Comparing Faces
• CYBORG: Blending Human Saliency Into the Loss Improves Deep Learning
• Semantic Image Fusion