# PCQA-R1: Advancing Generalized 3D Point Cloud Quality Assessment with Reinforcement Learning

arXiv: https://arxiv.org/abs/2608.18627 · HTML: https://mlsanity.com/paper/2608.18627/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 43 references that are in the map.
- First posted on arXiv 19 Aug 2026.

## Builds on (most cited first, 22 of 22)

- [Q-Align: Teaching LMMs for Visual Scoring via Discrete Text-Defined Levels](https://mlsanity.com/paper/2312.17090/) (2023)
- [Q-Insight: Understanding Image Quality via Visual Reinforcement Learning](https://mlsanity.com/paper/2503.22679/) (2025)
- [Inferring Point Cloud Quality via Graph Similarity](https://mlsanity.com/paper/2006.00497/) (2020)
- [Point Cloud Quality Assessment: Dataset Construction and Learning-based No-Reference Metric](https://mlsanity.com/paper/2012.11895/) (2020)
- [No-Reference Quality Assessment for 3D Colored Point Cloud and Mesh Models](https://mlsanity.com/paper/2107.02041/) (2021)
- [MM-PCQA: Multi-Modal Learning for No-reference Point Cloud Quality Assessment](https://mlsanity.com/paper/2209.00244/) (2022)
- [No-Reference Point Cloud Quality Assessment via Domain Adaptation](https://mlsanity.com/paper/2112.02851/) (2021)
- [Quality Assessment in the Era of Large Models: A Survey](https://mlsanity.com/paper/2409.00031/) (2024)
- [LMM-PCQA: Assisting Point Cloud Quality Assessment with LMM](https://mlsanity.com/paper/2404.18203/) (2024)
- [AGHI-QA: A Subjective-Aligned Dataset and Metric for AI-Generated Human Images](https://mlsanity.com/paper/2504.21308/) (2025)
- [PointPCA: Point Cloud Objective Quality Assessment Using PCA-Based Descriptors](https://mlsanity.com/paper/2111.12663/) (2021)
- [SingingHead: A Large-scale 4D Dataset for Singing Head Animation](https://mlsanity.com/paper/2312.04369/) (2023)
- [CLIP-PCQA: Exploring Subjective-Aligned Vision-Language Modeling for Point Cloud Quality Assessment](https://mlsanity.com/paper/2501.10071/) (2025)
- [PIT-QMM: A Large Multimodal Model For No-Reference Point Cloud Quality Assessment](https://mlsanity.com/paper/2510.07636/) (2025)
- [Asynchronous Feedback Network for Perceptual Point Cloud Quality Assessment](https://mlsanity.com/paper/2407.09806/) (2024)
- [QualiRAG: Retrieval-Augmented Generation for Visual Quality Understanding](https://mlsanity.com/paper/2601.18195/) (2026)
- [Q-Bench-Portrait: Benchmarking Multimodal Large Language Models on Portrait Image Quality Perception](https://mlsanity.com/paper/2601.18346/) (2026)
- [QD-PCQA: Quality-Aware Domain Adaptation for Point Cloud Quality Assessment](https://mlsanity.com/paper/2603.03726/) (2026)
- [Multi-Dimensional Quality Assessment for AI-Generated Human-Centric Videos: Dataset and Model](https://mlsanity.com/paper/2607.16742/) (2026)
- [From Images to Point Clouds: An Efficient Solution for Cross-media Blind Quality Assessment without Annotated Training](https://mlsanity.com/paper/2501.13387/) (2025)
- [UPDA: Unsupervised Progressive Domain Adaptation for No-Reference Point Cloud Quality Assessment](https://mlsanity.com/paper/2602.11969/) (2026)
- [FMReward: Aligning and Evaluating Audio-Driven 3D Facial Animation with Human Preferences](https://mlsanity.com/paper/2608.15296/) (2026)

## Abstract (arXiv)

No-reference point cloud quality assessment (PCQA) has been an active topic in recent years and is used to measure and optimize the visual experience of point clouds. However, large multimodal models (LMMs) have rarely been explored in this area. Previous LMM-based methods mainly rely on supervised fine-tuning to directly predict numerical quality scores, lacking the ability to generalize across datasets with heterogeneous MOS scales and limited annotations. A key difficulty is that absolute MOS regression can be brittle across datasets with different score scales and distortion distributions, whereas relative quality ranking is more stable under such shifts. In this paper, we present PCQA-R1, the first reinforcement learning LMM for 3D point cloud quality assessment to simultaneously model quality understanding and scoring. Built upon the group relative policy optimization (GRPO) strategy, PCQA-R1 first constructs a chain-of-thought dataset, PCQA-CoT, which serves as cold-start training data through a reverse reasoning strategy that teaches the LMM to generate its reasoning process. We further introduce a Gaussian proximity reward that prevents calibration drift by anchoring score predictions to the source MOS range. Experimental results demonstrate that PCQA-R1 achieves state-of-the-art cross-dataset generalization across five benchmarks and competitive in-domain accuracy. Ablation studies support the role of ranking, Gaussian reward, and cold-start traces.