SAM Struggles in Concealed Scenes -- Empirical Study on Segment Anything

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ji, Ge-Peng, Fan, Deng-Ping, Xu, Peng, Cheng, Ming-Ming, Zhou, Bowen, Van Gool, Luc
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916366736424960
author Ji, Ge-Peng
Fan, Deng-Ping
Xu, Peng
Cheng, Ming-Ming
Zhou, Bowen
Van Gool, Luc
author_facet Ji, Ge-Peng
Fan, Deng-Ping
Xu, Peng
Cheng, Ming-Ming
Zhou, Bowen
Van Gool, Luc
contents Segmenting anything is a ground-breaking step toward artificial general intelligence, and the Segment Anything Model (SAM) greatly fosters the foundation models for computer vision. We could not be more excited to probe the performance traits of SAM. In particular, exploring situations in which SAM does not perform well is interesting. In this report, we choose three concealed scenes, i.e., camouflaged animals, industrial defects, and medical lesions, to evaluate SAM under unprompted settings. Our main observation is that SAM looks unskilled in concealed scenes.
format Preprint
id arxiv_https___arxiv_org_abs_2304_06022
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle SAM Struggles in Concealed Scenes -- Empirical Study on Segment Anything
Ji, Ge-Peng
Fan, Deng-Ping
Xu, Peng
Cheng, Ming-Ming
Zhou, Bowen
Van Gool, Luc
Computer Vision and Pattern Recognition
Segmenting anything is a ground-breaking step toward artificial general intelligence, and the Segment Anything Model (SAM) greatly fosters the foundation models for computer vision. We could not be more excited to probe the performance traits of SAM. In particular, exploring situations in which SAM does not perform well is interesting. In this report, we choose three concealed scenes, i.e., camouflaged animals, industrial defects, and medical lesions, to evaluate SAM under unprompted settings. Our main observation is that SAM looks unskilled in concealed scenes.
title SAM Struggles in Concealed Scenes -- Empirical Study on Segment Anything
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2304.06022