EvObj: Learning Evolving Object-centric Representations for 3D Instance Segmentation without Scene Supervision
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909039660630016 |
|---|---|
| author | Chen, Jiahao Zhang, Zihui Yang, Yafei Li, Jinxi Wei, Shenxing Sun, Zhixuan Yang, Bo |
| author_facet | Chen, Jiahao Zhang, Zihui Yang, Yafei Li, Jinxi Wei, Shenxing Sun, Zhixuan Yang, Bo |
| contents | We introduce EvObj for unsupervised 3D instance segmentation that bridges the geometric domain gap between synthetic pretraining data and real-world point clouds. Current methods suffer from structural discrepancies when transferring object priors from synthetic datasets (e.g., ShapeNet) to real scans (e.g., ScanNet), particularly due to morphological variations and occlusion artifacts. To address this, EvObj integrates two innovative modules: (1) An object discerning module that dynamically refines object candidates, enabling continuous adaptation of object priors to target domains; and (2) An object completion module that reconstructs partial geometries after discovering objects. We conduct extensive experiments on both real-world and synthetic datasets, demonstrating superior 3D object segmentation performance over all baselines while achieving state-of-the-art results. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_13152 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | EvObj: Learning Evolving Object-centric Representations for 3D Instance Segmentation without Scene Supervision Chen, Jiahao Zhang, Zihui Yang, Yafei Li, Jinxi Wei, Shenxing Sun, Zhixuan Yang, Bo Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Robotics We introduce EvObj for unsupervised 3D instance segmentation that bridges the geometric domain gap between synthetic pretraining data and real-world point clouds. Current methods suffer from structural discrepancies when transferring object priors from synthetic datasets (e.g., ShapeNet) to real scans (e.g., ScanNet), particularly due to morphological variations and occlusion artifacts. To address this, EvObj integrates two innovative modules: (1) An object discerning module that dynamically refines object candidates, enabling continuous adaptation of object priors to target domains; and (2) An object completion module that reconstructs partial geometries after discovering objects. We conduct extensive experiments on both real-world and synthetic datasets, demonstrating superior 3D object segmentation performance over all baselines while achieving state-of-the-art results. |
| title | EvObj: Learning Evolving Object-centric Representations for 3D Instance Segmentation without Scene Supervision |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Robotics |
| url | https://arxiv.org/abs/2605.13152 |