Effective SAM Combination for Open-Vocabulary Semantic Segmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lee, Minhyeok, Cho, Suhwan, Lee, Jungho, Yang, Sunghun, Choi, Heeseung, Kim, Ig-Jae, Lee, Sangyoun
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909556621180928
author Lee, Minhyeok
Cho, Suhwan
Lee, Jungho
Yang, Sunghun
Choi, Heeseung
Kim, Ig-Jae
Lee, Sangyoun
author_facet Lee, Minhyeok
Cho, Suhwan
Lee, Jungho
Yang, Sunghun
Choi, Heeseung
Kim, Ig-Jae
Lee, Sangyoun
contents Open-vocabulary semantic segmentation aims to assign pixel-level labels to images across an unlimited range of classes. Traditional methods address this by sequentially connecting a powerful mask proposal generator, such as the Segment Anything Model (SAM), with a pre-trained vision-language model like CLIP. But these two-stage approaches often suffer from high computational costs, memory inefficiencies. In this paper, we propose ESC-Net, a novel one-stage open-vocabulary segmentation model that leverages the SAM decoder blocks for class-agnostic segmentation within an efficient inference framework. By embedding pseudo prompts generated from image-text correlations into SAM's promptable segmentation framework, ESC-Net achieves refined spatial aggregation for accurate mask predictions. ESC-Net achieves superior performance on standard benchmarks, including ADE20K, PASCAL-VOC, and PASCAL-Context, outperforming prior methods in both efficiency and accuracy. Comprehensive ablation studies further demonstrate its robustness across challenging conditions.
format Preprint
id arxiv_https___arxiv_org_abs_2411_14723
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Effective SAM Combination for Open-Vocabulary Semantic Segmentation
Lee, Minhyeok
Cho, Suhwan
Lee, Jungho
Yang, Sunghun
Choi, Heeseung
Kim, Ig-Jae
Lee, Sangyoun
Computer Vision and Pattern Recognition
Open-vocabulary semantic segmentation aims to assign pixel-level labels to images across an unlimited range of classes. Traditional methods address this by sequentially connecting a powerful mask proposal generator, such as the Segment Anything Model (SAM), with a pre-trained vision-language model like CLIP. But these two-stage approaches often suffer from high computational costs, memory inefficiencies. In this paper, we propose ESC-Net, a novel one-stage open-vocabulary segmentation model that leverages the SAM decoder blocks for class-agnostic segmentation within an efficient inference framework. By embedding pseudo prompts generated from image-text correlations into SAM's promptable segmentation framework, ESC-Net achieves refined spatial aggregation for accurate mask predictions. ESC-Net achieves superior performance on standard benchmarks, including ADE20K, PASCAL-VOC, and PASCAL-Context, outperforming prior methods in both efficiency and accuracy. Comprehensive ablation studies further demonstrate its robustness across challenging conditions.
title Effective SAM Combination for Open-Vocabulary Semantic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2411.14723