Being Negative but Constructively: Lessons Learnt from Creating Better Visual Question Answering Datasets
Preprint
Abstract
Visual question answering (QA) has attracted a lot of attention lately, seen essentially as a form of (visual) Turing test that artificial intelligence should strive to achieve. In this paper, we study a crucial component of this task: how can we design good datasets for the task? We focus on the design of multiple-choice based datasets where the learner has to select the right answer from a set of candidate ones including the target (i.e. the correct one) and the decoys (i.e. the incorrect ones). Through careful analysis of the results attained by state-of-the-art learning models and human annotators on existing datasets, we show the design of the decoy answers has a significant impact on how and what the learning models learn from the datasets. In particular, the resulting learner can ignore the visual information, the question, or the both while still doing well on the task. Inspired by this, we propose automatic procedures to remedy such design deficiencies. We apply the procedures to re-construct decoy answers for two popular visual QA datasets as well as to create a new visual QA dataset from the Visual Genome project, resulting in the largest dataset for this task. Extensive empirical studies show that the design deficiencies have been alleviated in the remedied datasets and the performance on them is likely a more faithful indicator of the difference among learning models. The datasets are released and publicly available via http://www.teds.usc.edu/website_vqa/.Keywords
All Related Versions
This publication has 17 references indexed in Scilit:
- Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question AnsweringInternational Journal of Computer Vision, 2018
- Visual question answering: Datasets, algorithms, and future challengesComputer Vision and Image Understanding, 2017
- Visual question answering: A survey of methods and datasetsComputer Vision and Image Understanding, 2017
- Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image AnnotationsInternational Journal of Computer Vision, 2017
- Leveraging Visual Question Answering for Image-Caption Ranking2016
- Revisiting Visual Question Answering Baselines2016
- Deep Residual Learning for Image Recognition2016
- Yin and Yang: Balancing and Answering Binary Visual Questions2016
- Visual7W: Grounded Question Answering in Images2016
- Microsoft COCO: Common Objects in Context2014