Details Matter for Indoor Open-vocabulary 3D Instance Segmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jung, Sanghun, Zheng, Jingjing, Zhang, Ke, Qiao, Nan, Chen, Albert Y. C., Xia, Lu, Liu, Chi, Sun, Yuyin, Zeng, Xiao, Huang, Hsiang-Wei, Boots, Byron, Sun, Min, Kuo, Cheng-Hao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909712747855872
author Jung, Sanghun
Zheng, Jingjing
Zhang, Ke
Qiao, Nan
Chen, Albert Y. C.
Xia, Lu
Liu, Chi
Sun, Yuyin
Zeng, Xiao
Huang, Hsiang-Wei
Boots, Byron
Sun, Min
Kuo, Cheng-Hao
author_facet Jung, Sanghun
Zheng, Jingjing
Zhang, Ke
Qiao, Nan
Chen, Albert Y. C.
Xia, Lu
Liu, Chi
Sun, Yuyin
Zeng, Xiao
Huang, Hsiang-Wei
Boots, Byron
Sun, Min
Kuo, Cheng-Hao
contents Unlike closed-vocabulary 3D instance segmentation that is often trained end-to-end, open-vocabulary 3D instance segmentation (OV-3DIS) often leverages vision-language models (VLMs) to generate 3D instance proposals and classify them. While various concepts have been proposed from existing research, we observe that these individual concepts are not mutually exclusive but complementary. In this paper, we propose a new state-of-the-art solution for OV-3DIS by carefully designing a recipe to combine the concepts together and refining them to address key challenges. Our solution follows the two-stage scheme: 3D proposal generation and instance classification. We employ robust 3D tracking-based proposal aggregation to generate 3D proposals and remove overlapped or partial proposals by iterative merging/removal. For the classification stage, we replace the standard CLIP model with Alpha-CLIP, which incorporates object masks as an alpha channel to reduce background noise and obtain object-centric representation. Additionally, we introduce the standardized maximum similarity (SMS) score to normalize text-to-proposal similarity, effectively filtering out false positives and boosting precision. Our framework achieves state-of-the-art performance on ScanNet200 and S3DIS across all AP and AR metrics, even surpassing an end-to-end closed-vocabulary method.
format Preprint
id arxiv_https___arxiv_org_abs_2507_23134
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Details Matter for Indoor Open-vocabulary 3D Instance Segmentation
Jung, Sanghun
Zheng, Jingjing
Zhang, Ke
Qiao, Nan
Chen, Albert Y. C.
Xia, Lu
Liu, Chi
Sun, Yuyin
Zeng, Xiao
Huang, Hsiang-Wei
Boots, Byron
Sun, Min
Kuo, Cheng-Hao
Computer Vision and Pattern Recognition
Unlike closed-vocabulary 3D instance segmentation that is often trained end-to-end, open-vocabulary 3D instance segmentation (OV-3DIS) often leverages vision-language models (VLMs) to generate 3D instance proposals and classify them. While various concepts have been proposed from existing research, we observe that these individual concepts are not mutually exclusive but complementary. In this paper, we propose a new state-of-the-art solution for OV-3DIS by carefully designing a recipe to combine the concepts together and refining them to address key challenges. Our solution follows the two-stage scheme: 3D proposal generation and instance classification. We employ robust 3D tracking-based proposal aggregation to generate 3D proposals and remove overlapped or partial proposals by iterative merging/removal. For the classification stage, we replace the standard CLIP model with Alpha-CLIP, which incorporates object masks as an alpha channel to reduce background noise and obtain object-centric representation. Additionally, we introduce the standardized maximum similarity (SMS) score to normalize text-to-proposal similarity, effectively filtering out false positives and boosting precision. Our framework achieves state-of-the-art performance on ScanNet200 and S3DIS across all AP and AR metrics, even surpassing an end-to-end closed-vocabulary method.
title Details Matter for Indoor Open-vocabulary 3D Instance Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2507.23134