# Question-Conditioned Counterfactual Image Generation for VQA

Source: [Question-Conditioned Counterfactual Image Generation for VQA](<https://www.sota2.com/research/paper/68fa66b01e209131cea11ce1>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T21:56:21.324Z

Authors: Jingjing Pan, Yash Goyal, Stefan Lee

Published: 2019-11-14T19:37:33.000Z

## Abstract

While Visual Question Answering (VQA) models continue to push the state-of-the-art forward, they largely remain black-boxes - failing to provide insight into how or why an answer is generated. In this ongoing work, we propose addressing this shortcoming by learning to generate counterfactual images for a VQA model - i.e. given a question-image pair, we wish to generate a new image such that i) the VQA model outputs a different answer, ii) the new image is minimally different from the original, and iii) the new image is realistic. Our hope is that providing such counterfactual examples allows users to investigate and understand the VQA model's internal mechanisms.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/1911.06352>)

[PDF](<https://arxiv.org/pdf/1911.06352.pdf>)
