From Wrong To Right: A Recursive Approach Towards Vision-Language Explanation
This work addresses the problem of data-efficient explanation generation for vision-language models, which is incremental as it builds on pre-trained models with a novel recursive approach.
The paper tackles the challenge of generating insightful explanations for visual reasoning tasks with limited annotations by introducing ReVisE, a recursive algorithm that iteratively improves explanation quality until answers converge, outperforming previous methods with only 5% of human-annotated explanations and achieving up to a 4.2 and 1.3 increase in BLEU-1 scores on VCR and VQA-X datasets.
Addressing the challenge of adapting pre-trained vision-language models for generating insightful explanations for visual reasoning tasks with limited annotations, we present ReVisE: a $\textbf{Re}$cursive $\textbf{Vis}$ual $\textbf{E}$xplanation algorithm. Our method iteratively computes visual features (conditioned on the text input), an answer, and an explanation, to improve the explanation quality step by step until the answer converges. We find that this multi-step approach guides the model to correct its own answers and outperforms single-step explanation generation. Furthermore, explanations generated by ReVisE also serve as valuable annotations for few-shot self-training. Our approach outperforms previous methods while utilizing merely 5% of the human-annotated explanations across 10 metrics, demonstrating up to a 4.2 and 1.3 increase in BLEU-1 score on the VCR and VQA-X datasets, underscoring the efficacy and data-efficiency of our method.