Generative Interventions for Causal Learning

We introduce a framework for learning robust visual representations that generalize to new viewpoints, backgrounds, and scene contexts. Discriminative models often learn naturally occurring spurious correlations, which cause them to fail on images outside of the training distribution. In this paper,...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2021-03
Main Authors: Mao, Chengzhi, Cha, Augustine, Gupta, Amogh, Wang, Hao, Yang, Junfeng, Vondrick, Carl
Format: Article
Language:English
Subjects:
Datasets
Learning
Performance enhancement
Representations
Robustness
Visual discrimination
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We introduce a framework for learning robust visual representations that generalize to new viewpoints, backgrounds, and scene contexts. Discriminative models often learn naturally occurring spurious correlations, which cause them to fail on images outside of the training distribution. In this paper, we show that we can steer generative models to manufacture interventions on features caused by confounding factors. Experiments, visualizations, and theoretical results show this method learns robust representations more consistent with the underlying causal relationships. Our approach improves performance on multiple datasets demanding out-of-distribution generalization, and we demonstrate state-of-the-art performance generalizing from ImageNet to ObjectNet dataset.
ISSN:2331-8422