Explainability-aided Domain Generalization for Image Classification

Abstract

Traditionally, for most machine learning settings, gaining some degree of explainability that tries to give users more insights into how and why the network arrives at its predictions, restricts the underlying model and hinders performance to a certain degree. For example, decision trees are thought of as being more explainable than deep neural networks but they lack performance on visual tasks. In this work, we empirically demonstrate that applying methods and architectures from the explainability literature can, in fact, achieve state-of-the-art performance for the challenging task of domain generalization while offering a framework for more insights into the prediction and training process. For that, we develop a set of novel algorithms including DivCAM, an approach where the network receives guidance during training via gradient based class activation maps to focus on a diverse set of discriminative features, as well as ProDrop and D-Transformers which apply prototypical networks to the domain generalization task, either with self-challenging or attention alignment. Since these methods offer competitive performance on top of explainability, we argue that the proposed methods can be used as a tool to improve the robustness of deep neural network architectures.

References in corpus (43)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Auto-Encoding Variational Bayes
• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Prototypical Networks for Few-shot Learning
• Axiomatic Attribution for Deep Networks
• Improved Regularization of Convolutional Neural Networks with Cutout
• Learning Important Features Through Propagating Activation Differences
• Equality of Opportunity in Supervised Learning
• Explaining NonLinear Classification Decisions with Deep Taylor Decomposition
• Theoretical Models of Learning to Learn
• Domain Separation Networks
• Domain Generalization via Invariant Feature Representation
• DropBlock: A regularization method for convolutional networks
• Attention is not Explanation
• Understanding Neural Networks through Representation Erasure
• Domain Generalization via Model-Agnostic Learning of Semantic Features
• Towards Robust Interpretability with Self-Explaining Neural Networks
• Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization
• On Fairness and Calibration
• e-SNLI: Natural Language Inference with Natural Language Explanations
• Out-of-Distribution Generalization via Risk Extrapolation (REx)
• Prototype selection for interpretable classification
• Deep Learning for Case-Based Reasoning through Prototypes: A Neural Network that Explains Its Predictions
• Recurrent Neural Networks (RNNs): A gentle Introduction and Overview
• Domain Generalization by Marginal Transfer Learning
• Improve Unsupervised Domain Adaptation with Mixup Training
• Heterogeneous Domain Generalization via Domain Mixup
• Understanding intermediate layers using linear classifier probes
• Does Distributionally Robust Supervised Learning Give Robust Classifiers?
• Learning Robust Representations by Projecting Superficial Statistics Out
• Generalizing to unseen domains via distribution matching
• Domain Generalization using Causal Matching
• Feature-Critic Networks for Heterogeneous Domain Generalization
• Frustratingly Simple Domain Generalization via Image Stylization
• Causal Learning and Explanation of Deep Neural Networks via Autoencoded Activations
• Self-Challenging Improves Cross-Domain Generalization
• When Unseen Domain Generalization is Unnecessary? Rethinking Data Augmentation
• Feature Alignment and Restoration for Domain Generalization and Adaptation
• Descending through a Crowded Valley - Benchmarking Deep Learning Optimizers
• Towards Shape Biased Unsupervised Representation Learning for Domain Generalization
• A Generalization Error Bound for Multi-class Domain Generalization
• Representation via Representations: Domain Generalization via Adversarially Learned Invariant Representations
• Prototypal Analysis and Prototypal Regression