Grounding Visual Explanations

TL;DR

This paper introduces a phrase-critic model that refines visual explanations by grounding them in the image, improving trustworthiness and enabling counterfactual reasoning for fine-grained classification.

Contribution

It proposes a novel phrase-critic approach that refines explanations with grounded phrases and detects incorrect explanations, enhancing visual explanation quality and reliability.

Findings

Improves explanation grounding in the CUB dataset.

Enhances detection and correction of incorrect explanations in FOIL tasks.

Provides counterfactual explanations for classification decisions.

Abstract

Existing visual explanation generating agents learn to fluently justify a class prediction. However, they may mention visual attributes which reflect a strong class prior, although the evidence may not actually be in the image. This is particularly concerning as ultimately such agents fail in building trust with human users. To overcome this limitation, we propose a phrase-critic model to refine generated candidate explanations augmented with flipped phrases which we use as negative examples while training. At inference time, our phrase-critic model takes an image and a candidate explanation as input and outputs a score indicating how well the candidate explanation is grounded in the image. Our explainable AI agent is capable of providing counter arguments for an alternative prediction, i.e. counterfactuals, along with explanations that justify the correct classification decisions. Our model improves the textual explanation quality of fine-grained classification decisions on the CUB dataset by mentioning phrases that are grounded in the image. Moreover, on the FOIL tasks, our agent detects when there is a mistake in the sentence, grounds the incorrect phrase and corrects it significantly better than other models.