Visual Explanation by Interpretation: Improving Visual Feedback Capabilities of Deep Neural Networks

Abstract

Interpretation and explanation of deep models is critical towards wide adoption of systems that rely on them. In this paper, we propose a novel scheme for both interpretation as well as explanation in which, given a pretrained model, we automatically identify internal features relevant for the set of classes considered by the model, without relying on additional annotations. We interpret the model through average visualizations of this reduced set of features. Then, at test time, we explain the network prediction by accompanying the predicted class label with supporting visualizations derived from the identified features. In addition, we propose a method to address the artifacts introduced by stridded operations in deconvNet-based visualizations. Moreover, we introduce an8Flower, a dataset specifically designed for objective quantitative evaluation of methods for visual explanation.Experiments on the MNIST,ILSVRC12,Fashion144k and an8Flower datasets show that our method produces detailed explanations with good coverage of relevant features of the classes of interest

Accepted at International Conference on Learning Representations (ICLR) 2019. Project website: http://homes.esat.kuleuven.be/~joramas/projects/visualExplanationByInterpretation

References in corpus (4)

• Grad-CAM++: Improved Visual Explanations for Deep Convolutional Networks
• Understanding Neural Networks Through Deep Visualization
• A Theoretical Explanation for Perplexing Behaviors of Backpropagation-based Visualizations
• Relating Input Concepts to Convolutional Neural Network Decisions

Cited by in corpus (7)

• Ground Truth Evaluation of Neural Network Explanations with CLEVR-XAI
• Explainable Artificial Intelligence: a Systematic Review
• Learning to synthesise the ageing brain without longitudinal data
• Embedding Deep Networks into Visual Explanations
• On The Coherence of Quantitative Evaluation of Visual Explanations
• Convolutional Neural Network Interpretability with General Pattern Theory
• MinMaxCAM: Improving object coverage for CAM-basedWeakly Supervised Object Localization