Saved in:
Bibliographic Details
Main Authors: Wang, Qizhou, Lin, Yong, Chen, Yongqiang, Schmidt, Ludwig, Han, Bo, Zhang, Tong
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2403.11497
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916465328783360
author Wang, Qizhou
Lin, Yong
Chen, Yongqiang
Schmidt, Ludwig
Han, Bo
Zhang, Tong
author_facet Wang, Qizhou
Lin, Yong
Chen, Yongqiang
Schmidt, Ludwig
Han, Bo
Zhang, Tong
contents Large vision language models, such as CLIP, demonstrate impressive robustness to spurious features than single-modal models trained on ImageNet. However, existing test datasets are typically curated based on ImageNet-trained models, which aim to capture the spurious features inherited in ImageNet. Benchmarking CLIP models based on the ImageNet-oriented spurious features may not be sufficient to reflect the extent to which CLIP models are robust to spurious correlations within CLIP training data, e.g., LAION. To this end, we craft a new challenging dataset named CounterAnimal designed to reveal the reliance of CLIP models on realistic spurious features. Specifically, we split animal photos into groups according to the backgrounds, and then identify a pair of groups for each class where a CLIP model shows high-performance drops across the two groups. Our evaluations show that the spurious features captured by CounterAnimal are generically learned by CLIP models with different backbones and pre-train data, yet have limited influence for ImageNet models. We provide theoretical insights that the CLIP objective cannot offer additional robustness. Furthermore, we also re-evaluate strategies such as scaling up parameters and high-quality pre-trained data. We find that they still help mitigate the spurious features, providing a promising path for future developments.
format Preprint
id arxiv_https___arxiv_org_abs_2403_11497
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Sober Look at the Robustness of CLIPs to Spurious Features
Wang, Qizhou
Lin, Yong
Chen, Yongqiang
Schmidt, Ludwig
Han, Bo
Zhang, Tong
Computer Vision and Pattern Recognition
Machine Learning
Large vision language models, such as CLIP, demonstrate impressive robustness to spurious features than single-modal models trained on ImageNet. However, existing test datasets are typically curated based on ImageNet-trained models, which aim to capture the spurious features inherited in ImageNet. Benchmarking CLIP models based on the ImageNet-oriented spurious features may not be sufficient to reflect the extent to which CLIP models are robust to spurious correlations within CLIP training data, e.g., LAION. To this end, we craft a new challenging dataset named CounterAnimal designed to reveal the reliance of CLIP models on realistic spurious features. Specifically, we split animal photos into groups according to the backgrounds, and then identify a pair of groups for each class where a CLIP model shows high-performance drops across the two groups. Our evaluations show that the spurious features captured by CounterAnimal are generically learned by CLIP models with different backbones and pre-train data, yet have limited influence for ImageNet models. We provide theoretical insights that the CLIP objective cannot offer additional robustness. Furthermore, we also re-evaluate strategies such as scaling up parameters and high-quality pre-trained data. We find that they still help mitigate the spurious features, providing a promising path for future developments.
title A Sober Look at the Robustness of CLIPs to Spurious Features
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2403.11497