Get Started
Home
Topics
Search
Library
Research questionHow can 3D class-incremental models learn new categories while remaining robust across heterogeneous point-cloud domains?As 3D models learn new object categories over time, point clouds from CAD models, scans, reconstructions, and corrupted observations can respond differently to continual updates. Standard catastrophic-forgetting measures may miss this domain-specific performance discrepancy.
AI
Computer Vision
Machine Learning
Research Paper
Technology
Latest papersRecent research connected to this question, newest first.Mitigating Performance Discrepancy in Cross-Domain 3D Class-Incremental LearningThe source studies heterogeneous-domain 3D class-incremental learning using clean CAD data, RGB-D scans, video reconstructions, and corrupted observations. It introduces the Domain3D-CIL protocol, adapts mainstream class-incremental learning baselines, and reports experiments with the exemplar-free PolyMem approach; the evidence is limited to the described benchmark evaluations.research paper · Sep 4, 2026
Related questions
How can monocular 3D reconstruction generalize across unseen object categories and arbitrary viewpoints?How can fine-grained audio-visual segmentation learn new classes continually without semantic drift or co-occurrence confusion?How can existing camera–LiDAR BEV detectors stay reliable under domain shifts and sensor failures without retraining?How can cross-platform LiDAR adaptation maintain accurate 3D localization despite viewpoint-induced pseudo-object point inconsistencies?