Loss re-scaling VQA: Revisiting the LanguagePrior Problem from a Class-imbalance View

TL;DR

This paper interprets the language prior problem in VQA as a class-imbalance issue and proposes a loss re-scaling method to improve model performance by addressing answer frequency disparities.

Contribution

It introduces a novel class-imbalance perspective for understanding the language prior problem and develops a loss re-scaling technique to mitigate answer frequency bias in VQA.

Findings

Improved accuracy on VQA-CP datasets using the proposed method.

The class imbalance interpretation is valid for other vision tasks.

The approach effectively reduces bias towards frequent answers.

Abstract

Recent studies have pointed out that many well-developed Visual Question Answering (VQA) models are heavily affected by the language prior problem, which refers to making predictions based on the co-occurrence pattern between textual questions and answers instead of reasoning visual contents. To tackle it, most existing methods focus on enhancing visual feature learning to reduce this superficial textual shortcut influence on VQA model decisions. However, limited effort has been devoted to providing an explicit interpretation for its inherent cause. It thus lacks a good guidance for the research community to move forward in a purposeful way, resulting in model construction perplexity in overcoming this non-trivial problem. In this paper, we propose to interpret the language prior problem in VQA from a class-imbalance view. Concretely, we design a novel interpretation scheme whereby the loss of mis-predicted frequent and sparse answers of the same question type is distinctly exhibited during the late training phase. It explicitly reveals why the VQA model tends to produce a frequent yet obviously wrong answer, to a given question whose right answer is sparse in the training set. Based upon this observation, we further develop a novel loss re-scaling approach to assign different weights to each answer based on the training data statistics for computing the final loss. We apply our approach into three baselines and the experimental results on two VQA-CP benchmark datasets evidently demonstrate its effectiveness. In addition, we also justify the validity of the class imbalance interpretation scheme on other computer vision tasks, such as face recognition and image classification.