Improving the trustworthiness of image classification models by utilizing bounding-box annotations

Abstract

We study utilizing auxiliary information in training data to improve the trustworthiness of machine learning models. Specifically, in the context of image classification, we propose to optimize a training objective that incorporates bounding box information, which is available in many image classification datasets. Preliminary experimental results show that the proposed algorithm achieves better performance in accuracy, robustness, and interpretability compared with baselines.

arXiv comment6 pages including references

References in corpus (5)

• Robustness May Be at Odds with Accuracy
• Unifying distillation and privileged information
• Hierarchical interpretations for neural network predictions
• Interpretations are useful: penalizing explanations to align neural networks with prior knowledge
• Bridging Adversarial Robustness and Gradient Interpretability