Saved in:
Bibliographic Details
Main Authors: Lei, Chenyang, Chen, Liyi, Cen, Jun, Chen, Xiao, Lei, Zhen, Heide, Felix, Liu, Ziwei, Chen, Qifeng, Zhang, Zhaoxiang
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2409.08083
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912024900927488
author Lei, Chenyang
Chen, Liyi
Cen, Jun
Chen, Xiao
Lei, Zhen
Heide, Felix
Liu, Ziwei
Chen, Qifeng
Zhang, Zhaoxiang
author_facet Lei, Chenyang
Chen, Liyi
Cen, Jun
Chen, Xiao
Lei, Zhen
Heide, Felix
Liu, Ziwei
Chen, Qifeng
Zhang, Zhaoxiang
contents Foundation models like ChatGPT and Sora that are trained on a huge scale of data have made a revolutionary social impact. However, it is extremely challenging for sensors in many different fields to collect similar scales of natural images to train strong foundation models. To this end, this work presents a simple and effective framework SimMAT to study an open problem: the transferability from vision foundation models trained on natural RGB images to other image modalities of different physical properties (e.g., polarization). SimMAT consists of a modality-agnostic transfer layer (MAT) and a pretrained foundation model. We apply SimMAT to a representative vision foundation model Segment Anything Model (SAM) to support any evaluated new image modality. Given the absence of relevant benchmarks, we construct a new benchmark to evaluate the transfer learning performance. Our experiments confirm the intriguing potential of transferring vision foundation models in enhancing other sensors' performance. Specifically, SimMAT can improve the segmentation performance (mIoU) from 22.15% to 53.88% on average for evaluated modalities and consistently outperforms other baselines. We hope that SimMAT can raise awareness of cross-modal transfer learning and benefit various fields for better results with vision foundation models.
format Preprint
id arxiv_https___arxiv_org_abs_2409_08083
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SimMAT: Exploring Transferability from Vision Foundation Models to Any Image Modality
Lei, Chenyang
Chen, Liyi
Cen, Jun
Chen, Xiao
Lei, Zhen
Heide, Felix
Liu, Ziwei
Chen, Qifeng
Zhang, Zhaoxiang
Computer Vision and Pattern Recognition
Foundation models like ChatGPT and Sora that are trained on a huge scale of data have made a revolutionary social impact. However, it is extremely challenging for sensors in many different fields to collect similar scales of natural images to train strong foundation models. To this end, this work presents a simple and effective framework SimMAT to study an open problem: the transferability from vision foundation models trained on natural RGB images to other image modalities of different physical properties (e.g., polarization). SimMAT consists of a modality-agnostic transfer layer (MAT) and a pretrained foundation model. We apply SimMAT to a representative vision foundation model Segment Anything Model (SAM) to support any evaluated new image modality. Given the absence of relevant benchmarks, we construct a new benchmark to evaluate the transfer learning performance. Our experiments confirm the intriguing potential of transferring vision foundation models in enhancing other sensors' performance. Specifically, SimMAT can improve the segmentation performance (mIoU) from 22.15% to 53.88% on average for evaluated modalities and consistently outperforms other baselines. We hope that SimMAT can raise awareness of cross-modal transfer learning and benefit various fields for better results with vision foundation models.
title SimMAT: Exploring Transferability from Vision Foundation Models to Any Image Modality
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2409.08083