Abstract

"Thinking in pictures," [1] i.e., spatial-temporal reasoning, effortless and instantaneous for humans, is believed to be a significant ability to perform logical induction and a crucial factor in the intellectual history of technology development. Modern Artificial Intelligence (AI), fueled by massive datasets, deeper models, and mighty computation, has come to a stage where (super-)human-level performances are observed in certain specific tasks. However, current AI's ability in "thinking in pictures" is still far lacking behind. In this work, we study how to improve machines' reasoning ability on one challenging task of this kind: Raven's Progressive Matrices (RPM). Specifically, we borrow the very idea of "contrast effects" from the field of psychology, cognition, and education to design and train a permutation-invariant model. Inspired by cognitive studies, we equip our model with a simple inference module that is jointly trained with the perception backbone. Combining all the elements, we propose the Contrastive Perceptual Inference network (CoPINet) and empirically demonstrate that CoPINet sets the new state-of-the-art for permutation-invariant models on two major datasets. We conclude that spatial-temporal reasoning depends on envisaging the possibilities consistent with the relations between objects and can be solved from pixel-level inputs.

Paper

Learning Perceptual Inference by Contrasting
Chi Zhang*, Baoxiong Jia*, Feng Gao, Yixin Zhu, Hongjing Lu, Song-Chun Zhu
Proceedings of Advances in Neural Information Processing Systems (NeurIPS), 2019
Spotlight (2.43% acceptance rate)
(* indicates equal contribution.)
Paper

Team

Chi Zhang1,4

Baoxiong Jia1

Feng Gao3,4

Yixin Zhu3,4

Hongjing Lu2

Song-Chun Zhu1,3,4

1 Department of Computer Science, UCLA

2 Department of Psychology, UCLA

3 Department of Statistics, UCLA

4 International Center for AI and Robot Autonomy (CARA)

Code

Coming soon.

Bibtex

@inproceedings{zhang2019learning,
author={Zhang, Chi and Jia, Baoxiong and Gao, Feng and Zhu, Yixin and Lu, Hongjing and Zhu, Song-Chun},
title={Learning Perceptual Inference by Contrasting},
booktitle={Proceedings of Advances in Neural Information Processing Systems (NeurIPS)},
year={2019}}