Research map / Point clouds / Point cloud classification
Dynamic 3D Point Cloud Sequences as 2D Videos
Yiming Zeng, Junhui Hou, Qijian Zhang, Siyu Ren, Wenping Wang
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 |
|---|---|---|---|
| replace point-based spatio-temporal encoder with SPCV 2D-video encoder | PST-Transformer · DeformingThings4D action recognition | accuracy (%): 76.19 -> 90.48 | Improved |
| use SPCV representation instead of point grouping for action recognition | PST-Transformer · DeformingThings4D action recognition | forward time (s): 0.010 -> 0.006 | Improved |
| use SPCV representation for temporal interpolation | IDEA-Net · DHB Swing | EMD (x10^3): 7.07 -> 2.74 | Improved |
| use SPCV representation for temporal interpolation | IDEA-Net · DHB Longdress | EMD (x10^3): 5.92 -> 2.16 | Improved |
| use SPCV representation for temporal interpolation | IDEA-Net · DHB | forward time (s): 0.106 -> 0.014 | Improved |
| use SPCV representation for temporal interpolation | IDEA-Net · DHB Longdress | CD (x10^3): 0.88 -> 0.85 | No clear effect |
| frame-wise structurization of static clouds versus RegGeoNet | RegGeoNet · Dancer | CD (x10^-4): 0.659 -> 0.034 | Improved |
| frame-wise structurization of static clouds versus Flattening-Net | Flattening-Net · Chair | CD (x10^-4): 2.881 -> 0.086 | Improved |
| add sequence-wise structurization (deformation fields) (ablation) | SPCV · point cloud sequences | temporal consistency | Improved |
| add spatial regularization (lambda_s > 0) (ablation) | SPCV · point cloud sequences | spatial smoothness | Improved |
| add normal regularization (lambda_n > 0) (ablation) | SPCV · point cloud sequences | spatial smoothness and temporal consistency | Improved |
| use the paper's default distance instead of EMD (ablation) | SPCV · point cloud sequences | smoothness, consistency, fidelity | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- IDEA-Net — IDEA-Net: Dynamic 3D Point Cloud Interpolation via Deep Embedding Alignment (CVPR 2022) · better on 4 of 4 shared results
- PSTNet2 — PSTNet: Point Spatio-Temporal Convolution on Point Cloud Sequences (ICLR 2021) · better on 3 of 3 shared results
- PointINet — PointINet: Point Cloud Frame Interpolation Network (AAAI 2021) · 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
- 3DInAction: Understanding Human Actions in 3D Point Clouds (CVPR 2024 · similarity 0.88)
- Adapting Pre-trained 3D Models for Point Cloud Video Understanding via Cross-frame Spatio-temporal Perception (CVPR 2025 · similarity 0.88)
- Fast Encoder-Based 3D from Casual Videos via Point Track Processing (NeurIPS 2024 · similarity 0.87)
- PvNeXt: Rethinking Network Design and Temporal Motion for Point Cloud Video Recognition (ICLR 2025 · similarity 0.87)
- Point2Real: Bridging the Gap between Point Cloud and Realistic Image for Open-World 3D Recognition (AAAI 2024 · similarity 0.87)
- Stereo4D: Learning How Things Move in 3D from Internet Stereo Videos (CVPR 2025 · similarity 0.87)
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.
