Research map / Point clouds / Point cloud classification
Point 4D Transformer Networks for Spatio-Temporal Modeling in Point Cloud Videos
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use P4Transformer instead of MeteorNet | point-cloud action recognition · MSR-Action3D, 24 frames | accuracy: 88.50 -> 90.94 | Helped |
| use P4Transformer instead of 3DV-PointNet++ | 3D action recognition · NTU RGB+D 120, cross-subject | accuracy: 82.4 -> 86.4 | Helped |
| use P4Transformer instead of MeteorNet-l | 4D semantic segmentation · Synthia 4D, 3 frames | mIoU: 81.80 -> 83.16 | Helped |
| increase input clip length from 4 to 24 frames | P4Transformer · MSR-Action3D | accuracy: 80.13 -> 90.94 | Helped |
| use frame-level instead of video-level self-attention (ablation) | P4Transformer · MSR-Action3D, 12 frames | accuracy: 87.54 -> 70.65 | Hurt |
| use frame-level instead of video-level self-attention (ablation) | P4Transformer · MSR-Action3D, 16 frames | accuracy: 89.56 -> 76.45 | Hurt |
| use frame-level instead of video-level self-attention (ablation) | P4Transformer · MSR-Action3D, 20 frames | accuracy: 90.24 -> 78.16 | Hurt |
| use frame-level instead of video-level self-attention (ablation) | P4Transformer · MSR-Action3D, 24 frames | accuracy: 90.94 -> 79.18 | Hurt |
| increase temporal radius beyond 1 (ablation) | P4Transformer point 4D convolution · MSR-Action3D | accuracy | Hurt |
| increase spatial radius from too small to a suitable radius (ablation) | P4Transformer point 4D convolution · MSR-Action3D | accuracy | Helped |
| use an excessively large spatial radius (ablation) | P4Transformer point 4D convolution · MSR-Action3D | accuracy | Hurt |
| stack more transformer layers, up to a suitable depth (ablation) | P4Transformer · MSR-Action3D | accuracy | Helped |
Closest papers by meaning
- X4D-SceneFormer: Enhanced Scene Understanding on 4D Point Cloud Videos through Cross-Modal Knowledge Transfer (AAAI 2024 · similarity 0.90)
- Adapting Pre-trained 3D Models for Point Cloud Video Understanding via Cross-frame Spatio-temporal Perception (CVPR 2025 · similarity 0.88)
- 3DInAction: Understanding Human Actions in 3D Point Clouds (CVPR 2024 · similarity 0.87)
- Point Transformer V3: Simpler Faster Stronger (CVPR 2024 · similarity 0.86)
- PvNeXt: Rethinking Network Design and Temporal Motion for Point Cloud Video Recognition (ICLR 2025 · similarity 0.86)
- Mamba4D: Efficient 4D Point Cloud Video Understanding with Disentangled Spatial-Temporal State Space Models (CVPR 2025 · similarity 0.85)
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.