Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2410.10817 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917802234871808 |
|---|---|
| author | Sundaram, Shobhita Fu, Stephanie Muttenthaler, Lukas Tamir, Netanel Y. Chai, Lucy Kornblith, Simon Darrell, Trevor Isola, Phillip |
| author_facet | Sundaram, Shobhita Fu, Stephanie Muttenthaler, Lukas Tamir, Netanel Y. Chai, Lucy Kornblith, Simon Darrell, Trevor Isola, Phillip |
| contents | Humans judge perceptual similarity according to diverse visual attributes, including scene layout, subject location, and camera pose. Existing vision models understand a wide range of semantic abstractions but improperly weigh these attributes and thus make inferences misaligned with human perception. While vision representations have previously benefited from alignment in contexts like image generation, the utility of perceptually aligned representations in more general-purpose settings remains unclear. Here, we investigate how aligning vision model representations to human perceptual judgments impacts their usability across diverse computer vision tasks. We finetune state-of-the-art models on human similarity judgments for image triplets and evaluate them across standard vision benchmarks. We find that aligning models to perceptual judgments yields representations that improve upon the original backbones across many downstream tasks, including counting, segmentation, depth estimation, instance retrieval, and retrieval-augmented generation. In addition, we find that performance is widely preserved on other tasks, including specialized out-of-distribution domains such as in medical imaging and 3D environment frames. Our results suggest that injecting an inductive bias about human perceptual knowledge into vision models can contribute to better representations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_10817 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | When Does Perceptual Alignment Benefit Vision Representations? Sundaram, Shobhita Fu, Stephanie Muttenthaler, Lukas Tamir, Netanel Y. Chai, Lucy Kornblith, Simon Darrell, Trevor Isola, Phillip Computer Vision and Pattern Recognition Machine Learning Humans judge perceptual similarity according to diverse visual attributes, including scene layout, subject location, and camera pose. Existing vision models understand a wide range of semantic abstractions but improperly weigh these attributes and thus make inferences misaligned with human perception. While vision representations have previously benefited from alignment in contexts like image generation, the utility of perceptually aligned representations in more general-purpose settings remains unclear. Here, we investigate how aligning vision model representations to human perceptual judgments impacts their usability across diverse computer vision tasks. We finetune state-of-the-art models on human similarity judgments for image triplets and evaluate them across standard vision benchmarks. We find that aligning models to perceptual judgments yields representations that improve upon the original backbones across many downstream tasks, including counting, segmentation, depth estimation, instance retrieval, and retrieval-augmented generation. In addition, we find that performance is widely preserved on other tasks, including specialized out-of-distribution domains such as in medical imaging and 3D environment frames. Our results suggest that injecting an inductive bias about human perceptual knowledge into vision models can contribute to better representations. |
| title | When Does Perceptual Alignment Benefit Vision Representations? |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2410.10817 |