# Refine Then Fusion: Training-Free 3D Point Cloud Adaptation with Priority Refinement and Multi-Modal Knowledge Fusion

arXiv: https://arxiv.org/abs/2609.21522 · HTML: https://mlsanity.com/paper/2609.21522/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 38 references that are in the map.
- First posted on arXiv 18 Sep 2026.

## Builds on (most cited first, 21 of 21)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Zero-Shot Text-to-Image Generation](https://mlsanity.com/paper/2102.12092/) (2021)
- [LXMERT: Learning Cross-Modality Encoder Representations from Transformers](https://mlsanity.com/paper/1908.07490/) (2019)
- [CLIP-Adapter: Better Vision-Language Models with Feature Adapters](https://mlsanity.com/paper/2110.04544/) (2021)
- [Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World Data](https://mlsanity.com/paper/1908.04616/) (2019)
- [PointCLIP: Point Cloud Understanding by CLIP](https://mlsanity.com/paper/2112.02413/) (2021)
- [ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understanding](https://mlsanity.com/paper/2212.05171/) (2022)
- [LL3DA: Visual Interactive Instruction Tuning for Omni-3D Understanding, Reasoning, and Planning](https://mlsanity.com/paper/2311.18651/) (2023)
- [ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understanding](https://mlsanity.com/paper/2305.08275/) (2023)
- [PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learning](https://mlsanity.com/paper/2211.11682/) (2022)
- [OpenShape: Scaling Up 3D Shape Representation Towards Open-World Understanding](https://mlsanity.com/paper/2305.10764/) (2023)
- [Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining](https://mlsanity.com/paper/2302.02318/) (2023)
- [CLIP2Point: Transfer CLIP to Point Cloud Classification with Image-Depth Pre-training](https://mlsanity.com/paper/2210.01055/) (2022)
- [Not All Features Matter: Enhancing Few-shot CLIP with Adaptive Prior Refinement](https://mlsanity.com/paper/2304.01195/) (2023)
- [CLIP goes 3D: Leveraging Prompt Tuning for Language Grounded 3D Recognition](https://mlsanity.com/paper/2303.11313/) (2023)
- [MoPE-CLIP: Structured Pruning for Efficient Vision-Language Models with Module-wise Pruning Error Metric](https://mlsanity.com/paper/2403.07839/) (2024)
- [ViT-Lens: Towards Omni-modal Representations](https://mlsanity.com/paper/2311.16081/) (2023)
- [No Time to Train: Empowering Non-Parametric Networks for Few-shot 3D Scene Segmentation](https://mlsanity.com/paper/2404.04050/) (2024)
- [GAPrompt: Geometry-Aware Point Cloud Prompt for 3D Vision Model](https://mlsanity.com/paper/2505.04119/) (2025)
- [Beyond First Impressions: Integrating Joint Multi-modal Cues for Comprehensive 3D Representation](https://mlsanity.com/paper/2308.02982/) (2023)
- [Revisiting Few-Shot Learning from a Causal Perspective](https://mlsanity.com/paper/2209.13816/) (2022)

## Abstract (arXiv)

Recent pre-trained foundation models provide rich multi-modal priors for downstream 3D vision tasks. However, the effectiveness of these representations in few-shot scenarios is limited by two fundamental challenges: High-dimensional features often contain substantial channel redundancy and task-irrelevant noise, while the reliability of different modalities varies across samples. Consequently, direct aggregation of heterogeneous representations overlooks sample-dependent modality reliability and may obscure the discriminative cues essential. To address these limitations, we propose Refine Then Fusion(RTF), a training-free framework for few-shot 3D recognition. RTF first identifies discriminative feature channels by jointly modeling inter-class similarity and intra-class stability, thereby decoupling domain-specific knowledge refinement from the cached representations of pre-trained models. It then introduces a reliability-aware fusion mechanism that estimates sample-wise modality reliability from the distribution shifts induced by feature refinement, enabling adaptive aggregation of multi-modal representations. Furthermore, RTF constructs a memory cache that integrates instance-level support features with class-level prototypes to infer query labels. Extensive experiments on five benchmarks demonstrate that RTF consistently outperforms single-modal baselines, partial-fusion variants, and existing lightweight adaptation methods, achieving state-of-the-art few-shot 3D recognition performance without gradient optimization, additional training data, auxiliary training, or parameter updates.