English

LRTA: A Transparent Neural-Symbolic Reasoning Framework with Modular Supervision for Visual Question Answering

Computation and Language 2020-11-24 v1 Artificial Intelligence Computer Vision and Pattern Recognition Machine Learning

Abstract

The predominant approach to visual question answering (VQA) relies on encoding the image and question with a "black-box" neural encoder and decoding a single token as the answer like "yes" or "no". Despite this approach's strong quantitative results, it struggles to come up with intuitive, human-readable forms of justification for the prediction process. To address this insufficiency, we reformulate VQA as a full answer generation task, which requires the model to justify its predictions in natural language. We propose LRTA [Look, Read, Think, Answer], a transparent neural-symbolic reasoning framework for visual question answering that solves the problem step-by-step like humans and provides human-readable form of justification at each step. Specifically, LRTA learns to first convert an image into a scene graph and parse a question into multiple reasoning instructions. It then executes the reasoning instructions one at a time by traversing the scene graph using a recurrent neural-symbolic execution module. Finally, it generates a full answer to the given question with natural language justifications. Our experiments on GQA dataset show that LRTA outperforms the state-of-the-art model by a large margin (43.1% v.s. 28.0%) on the full answer generation task. We also create a perturbed GQA test set by removing linguistic cues (attributes and relations) in the questions for analyzing whether a model is having a smart guess with superficial data correlations. We show that LRTA makes a step towards truly understanding the question while the state-of-the-art model tends to learn superficial correlations from the training data.

Keywords

Cite

@article{arxiv.2011.10731,
  title  = {LRTA: A Transparent Neural-Symbolic Reasoning Framework with Modular Supervision for Visual Question Answering},
  author = {Weixin Liang and Feiyang Niu and Aishwarya Reganti and Govind Thattai and Gokhan Tur},
  journal= {arXiv preprint arXiv:2011.10731},
  year   = {2020}
}

Comments

NeurIPS KR2ML 2020

R2 v1 2026-06-23T20:24:38.759Z