Research map / Point clouds / Point cloud classification
OpenShape: Scaling Up 3D Shape Representation Towards Open-World Understanding
Minghua Liu, Ruoxi Shi, Kaiming Kuang, Yinhao Zhu, Xuanlin Li, Shizhong Han, Hong Cai, Fatih Porikli, Hao Su
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| train on the ensembled 4-dataset shapes instead of ShapeNet only (ablation) | OpenShape-PointBERT · Objaverse-LVIS zero-shot | top-1 accuracy (%): 10.8 -> 46.8 | Improved |
| OpenShape training recipe (text enrichment, backbone scaling) instead of ULIP on same ensembled data | PointBERT · Objaverse-LVIS zero-shot | top-1 accuracy (%): 26.8 -> 46.8 | Improved |
| include Objaverse-LVIS shapes in training (Ensembled vs no LVIS) | OpenShape-PointBERT · ModelNet40 zero-shot | top-1 accuracy (%): 85.3 -> 84.4 | Hurt |
| add text filtering and enrichment (captioning, retrieval) (ablation) | OpenShape SparseConv/PointBERT · Objaverse-LVIS and ModelNet40 zero-shot | top-1 accuracy | Improved |
| scale up backbone parameters (ablation) | 3D backbones on ensembled dataset · Objaverse-LVIS zero-shot | top-1 accuracy | Improved |
| scale PointBERT from 32.3M to 72.1M parameters (ablation) | PointBERT · all benchmarks | zero-shot accuracy | Hurt |
| add hard negative mining (ablation) | OpenShape-SparseConv · Objaverse-LVIS and ModelNet40 zero-shot | top-1 accuracy | Improved |
| fine-tune CLIP text encoder for one epoch (ablation) | OpenShape-PointBERT · benchmarks | zero-shot accuracy | No clear effect |
| remove RGB color, use xyz only (ablation) | OpenShape · zero-shot classification | top-1 accuracy | No clear effect |
| add both image captioning and text retrieval (ablation) | OpenShape · Objaverse-LVIS zero-shot | top-1 accuracy | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PointCLIP v2 — PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learning (ICCV 2023) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Uni3D: Exploring Unified 3D Representation at Scale (ICLR 2024 · similarity 0.89)
- Towards 3D Objectness Learning in an Open World (NeurIPS 2025 · similarity 0.88)
- OpenDlign: Open-World Point Cloud Understanding with Depth-Aligned Images (NeurIPS 2024 · similarity 0.88)
- Cross-Modal 3D Representation with Multi-View Images and Point Clouds (CVPR 2025 · similarity 0.88)
- ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding (CVPR 2024 · similarity 0.88)
- ImOV3D: Learning Open Vocabulary Point Clouds 3D Object Detection from Only 2D Images (NeurIPS 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
