• Home
  • Search
  • Improving Automatic VQA Evaluation Using Large Language Models
  • Open Access IconOpen Access
  • Cite Icon30
  • https://doi.org/10.1609/aaai.v38i5.28212Copy DOI Icon

Improving Automatic VQA Evaluation Using Large Language Models

Show More
  • Abstract
  • Literature Map
  • Citations
  • Similar Papers
Abstract

8 years after the visual question answering (VQA) task was proposed, accuracy remains the primary metric for automatic evaluation. VQA Accuracy has been effective so far in the IID evaluation setting. However, our community is undergoing a shift towards open-ended generative models and OOD evaluation. In this new paradigm, the existing VQA Accuracy metric is overly stringent and underestimates the performance of VQA systems. Thus, there is a need to develop more robust automatic VQA metrics that serve as a proxy for human judgment. In this work, we propose to leverage the in-context learning capabilities of instruction-tuned large language models (LLMs) to build a better VQA metric. We formulate VQA evaluation as an answer-rating task where the LLM is instructed to score the accuracy of a candidate answer given a set of reference answers. We demonstrate the proposed metric better correlates with human judgment compared to existing metrics across several VQA models and benchmarks. We hope wide adoption of our metric will contribute to better estimating the research progress on the VQA task. We plan to release the evaluation code and collected human judgments.

Similar Papers
  • Book Chapter
  • Citations5

Neural Networks for Detecting Irrelevant Questions During Visual Question Answering

  • Jan 01, 2020
  • Mengdi Li +2
  • Research Article
  • Citations20

A framework for visual question answering with the integration of scene-text using PHOCs and fisher vectors

  • Nov 07, 2021
  • Expert Systems with Applications
  • Himanshu Sharma +1
  • Research Article
  • Citations5

Co-LLaVA: Efficient Remote Sensing Visual Question Answering via Model Collaboration

  • Jan 29, 2025
  • Remote Sensing
  • Fan Liu +5
  • Research Article
  • Citations8

Prophet: Prompting Large Language Models With Complementary Answer Heuristics for Knowledge-Based Visual Question Answering.

  • Aug 01, 2025
  • IEEE transactions on pattern analysis and machine intelligence
  • Zhou Yu +4
  • Conference Article
  • Citations2558

Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering

  • Jul 01, 2017
  • Yash Goyal +4
  • Conference Article
  • Citations48

Coarse-to-Fine Reasoning for Visual Question Answering

  • Jun 01, 2022
  • Binh X Nguyen +5
  • Research Article
  • Citations17

Accuracy vs. complexity: A trade-off in visual question answering models

  • Jun 12, 2021
  • Pattern Recognition
  • Moshiur Farazi +2
  • Conference Article
  • Citations2

End-To-End Training Of Object Segmentation Task And Video Question-Answering Task

  • Jan 03, 2023
  • Hidemoto Nakada +1
  • Research Article
  • Citations15

MobiVQA

  • Jul 04, 2022
  • Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
  • Qingqing Cao +3
  • Conference Article
  • Citations11

The Impact of Explanations on AI Competency Prediction in VQA

  • Sep 01, 2020
  • Kamran Alipour +5
  • Research Article
  • Citations6

Explanation vs. attention: A two-player game to obtain attention for VQA and visual dialog

  • Jul 23, 2022
  • Pattern Recognition
  • Badri N Patro +2
  • Research Article
  • Citations29

Explanation vs Attention: A Two-Player Game to Obtain Attention for VQA

  • Apr 03, 2020
  • Proceedings of the AAAI Conference on Artificial Intelligence
  • Badri Patro +2
  • Research Article
  • Citations2

Hierarchical Attention Networks for Multimodal Machine Learning

  • Mar 01, 2022
  • Journal of Physics Conference Series
  • Haotian Liang +1
  • Research Article
  • Citations27

Toward Explainable 3D Grounded Visual Question Answering: A New Benchmark and Strong Baseline

  • Jun 01, 2023
  • IEEE Transactions on Circuits and Systems for Video Technology
  • Lichen Zhao +8
  • Research Article
  • Citations9

VQAttack: Transferable Adversarial Attacks on Visual Question Answering via Pre-trained Models

  • Mar 24, 2024
  • Proceedings of the AAAI Conference on Artificial Intelligence
  • Ziyi Yin +7
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.