RGB-D Video Object Segmentation via Enhanced Multi-store Feature Memory

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xu, Boyue, Hou, Ruichao, Ren, Tongwei, Wu, Gangshan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916703424741376
author Xu, Boyue
Hou, Ruichao
Ren, Tongwei
Wu, Gangshan
author_facet Xu, Boyue
Hou, Ruichao
Ren, Tongwei
Wu, Gangshan
contents The RGB-Depth (RGB-D) Video Object Segmentation (VOS) aims to integrate the fine-grained texture information of RGB with the spatial geometric clues of depth modality, boosting the performance of segmentation. However, off-the-shelf RGB-D segmentation methods fail to fully explore cross-modal information and suffer from object drift during long-term prediction. In this paper, we propose a novel RGB-D VOS method via multi-store feature memory for robust segmentation. Specifically, we design the hierarchical modality selection and fusion, which adaptively combines features from both modalities. Additionally, we develop a segmentation refinement module that effectively utilizes the Segmentation Anything Model (SAM) to refine the segmentation mask, ensuring more reliable results as memory to guide subsequent segmentation tasks. By leveraging spatio-temporal embedding and modality embedding, mixed prompts and fused images are fed into SAM to unleash its potential in RGB-D VOS. Experimental results show that the proposed method achieves state-of-the-art performance on the latest RGB-D VOS benchmark.
format Preprint
id arxiv_https___arxiv_org_abs_2504_16471
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle RGB-D Video Object Segmentation via Enhanced Multi-store Feature Memory
Xu, Boyue
Hou, Ruichao
Ren, Tongwei
Wu, Gangshan
Computer Vision and Pattern Recognition
The RGB-Depth (RGB-D) Video Object Segmentation (VOS) aims to integrate the fine-grained texture information of RGB with the spatial geometric clues of depth modality, boosting the performance of segmentation. However, off-the-shelf RGB-D segmentation methods fail to fully explore cross-modal information and suffer from object drift during long-term prediction. In this paper, we propose a novel RGB-D VOS method via multi-store feature memory for robust segmentation. Specifically, we design the hierarchical modality selection and fusion, which adaptively combines features from both modalities. Additionally, we develop a segmentation refinement module that effectively utilizes the Segmentation Anything Model (SAM) to refine the segmentation mask, ensuring more reliable results as memory to guide subsequent segmentation tasks. By leveraging spatio-temporal embedding and modality embedding, mixed prompts and fused images are fed into SAM to unleash its potential in RGB-D VOS. Experimental results show that the proposed method achieves state-of-the-art performance on the latest RGB-D VOS benchmark.
title RGB-D Video Object Segmentation via Enhanced Multi-store Feature Memory
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2504.16471