Why is [person4
] pointing at [person1
]?
Rationale: I think so becauseâŠ
Visual Commonsense Reasoning (VCR) is a new task and large-scale dataset for cognition-level visual understanding.
With one glance at an image, we can effortlessly imagine the world beyond the pixels (e.g. that [person1] ordered pancakes). While this task is easy for humans, it is tremendously difficult for todayâs vision systems, requiring higher-order cognition and commonsense reasoning about the world. We formalize this task as Visual Commonsense Reasoning. In addition to answering challenging visual questions expressed in natural language, a model must provide a rationale explaining why its answer is true.
Overview of VCR
- 290k multiple choice questions
- 290k correct answers and rationales: one per question
- 110k images
- Counterfactual choices obtained with minimal bias, via our new Adversarial Matching approach
- Answers are 7.5 words on average; rationales are 16 words.
- High human agreement (>90%)
- Scaffolded on top of 80 object categories from COCO
- Questions are highly diverse and challenging: browse and see for yourself!
From Recognition to Cognition: Visual Commonsense Reasoning
If the paper inspires you, please cite us:
@inproceedings{zellers2019vcr,
author = {Zellers, Rowan and Bisk, Yonatan and Farhadi, Ali and Choi, Yejin},
title = {From Recognition to Cognition: Visual Commonsense Reasoning},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2019}
}
Authors
VCR is an effort between researchers at the University of Washington and AI2, along with a group of fantastic crowd workers who annotated the data. Weâre also grateful for the following sponsors: