Research map / Point clouds / Point cloud classification
Points Meet Pixels: Bridging 2D Vision-Language Model and 3D Perception Gaps for Point Cloud Quality Assessment
Mingxuan Li, Zihao Huang, Xiaohui Chu, Fazhan Zhang, Bohan Fu, Runze Hu
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 |
|---|---|---|---|
| PMP-PCQA vs second-best method | LMM-PCQA · SJTU-PCQA | SRCC: 0.925 -> 0.957 | Improved |
| PMP-PCQA vs second-best method | LMM-PCQA · LS-PCQA | SRCC: 0.731 -> 0.763 | Improved |
| PMP-PCQA vs second-best method | CLIP-PCQA · BASICS | SRCC: 0.856 -> 0.892 | Improved |
| add FCA module (ablation) | PMP-PCQA · SJTU-PCQA | SRCC: 0.923 -> 0.957 | Improved |
| add SAE module (ablation) | PMP-PCQA · SJTU-PCQA | SRCC: 0.943 -> 0.957 | Improved |
| add TAM module (ablation) | PMP-PCQA · SJTU-PCQA | SRCC: 0.946 -> 0.957 | Improved |
| add FCA module (ablation) | PMP-PCQA · LS-PCQA | SRCC: 0.722 -> 0.763 | Improved |
| use entropy-adaptive threshold instead of fixed 0.6 (ablation) | TAM · SJTU-PCQA | SRCC: 0.947 -> 0.957 | Improved |
| use 64 sampling points instead of 256 (ablation) | SAE · SJTU-PCQA | SRCC: 0.941 -> 0.957 | Improved |
| use 64 sampling points instead of 16 (ablation) | SAE · SJTU-PCQA | SRCC: 0.945 -> 0.957 | Improved |
| add text branch (ablation) | image+point model · SJTU-PCQA | SRCC: 0.925 -> 0.957 | Improved |
| add image branch to point+text (ablation) | point+text model · SJTU-PCQA | SRCC: 0.839 -> 0.957 | 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.
- CLIP-PCQA — CLIP-PCQA: Exploring Subjective-Aligned Vision-Language Modeling for Point Cloud Quality Assessment (AAAI 2025) · better on 3 of 3 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
- CLIP-PCQA: Exploring Subjective-Aligned Vision-Language Modeling for Point Cloud Quality Assessment (AAAI 2025 · similarity 0.92)
- Learning Disentangled Representations for Perceptual Point Cloud Quality Assessment via Mutual Information Minimization (NeurIPS 2024 · similarity 0.88)
- Contrastive Pre-Training with Multi-View Fusion for No-Reference Point Cloud Quality Assessment (CVPR 2024 · similarity 0.88)
- SD-VLM: Spatial Measuring and Understanding with Depth-Encoded Vision-Language Models (NeurIPS 2025 · similarity 0.88)
- Geometrically-driven Aggregation for Zero-shot 3D Point Cloud Understanding (CVPR 2024 · similarity 0.87)
- ReVSI: Rebuilding Visual Spatial Intelligence Evaluation for Accurate Assessment of VLM 3D Reasoning (ICML 2026 · 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.
