English

WeaQA: Weak Supervision via Captions for Visual Question Answering

Computer Vision and Pattern Recognition 2021-05-31 v2 Computation and Language

Abstract

Methodologies for training visual question answering (VQA) models assume the availability of datasets with human-annotated \textit{Image-Question-Answer} (I-Q-A) triplets. This has led to heavy reliance on datasets and a lack of generalization to new types of questions and scenes. Linguistic priors along with biases and errors due to annotator subjectivity have been shown to percolate into VQA models trained on such samples. We study whether models can be trained without any human-annotated Q-A pairs, but only with images and their associated textual descriptions or captions. We present a method to train models with synthetic Q-A pairs generated procedurally from captions. Additionally, we demonstrate the efficacy of spatial-pyramid image patches as a simple but effective alternative to dense and costly object bounding box annotations used in existing VQA models. Our experiments on three VQA benchmarks demonstrate the efficacy of this weakly-supervised approach, especially on the VQA-CP challenge, which tests performance under changing linguistic priors.

Keywords

Cite

@article{arxiv.2012.02356,
  title  = {WeaQA: Weak Supervision via Captions for Visual Question Answering},
  author = {Pratyay Banerjee and Tejas Gokhale and Yezhou Yang and Chitta Baral},
  journal= {arXiv preprint arXiv:2012.02356},
  year   = {2021}
}

Comments

Accepted in Findings of ACL 2021

R2 v1 2026-06-23T20:43:24.990Z