Monte Carlo Diffusion for Generalizable Learning-Based RANSAC

2025-03-12

AI Summary

Existing learning-based RANSAC methods suffer from severely limited out-of-distribution (OOD) generalization due to their reliance on training and test data drawn from identical distributions. To address this, we propose Monte Carlo Diffusion RANSAC (MCD-RANSAC), the first framework integrating diffusion modeling with multi-stage Monte Carlo sampling. By injecting controllable noise to explicitly model real-world observation uncertainty, MCD-RANSAC decouples model training from dependence on any specific data distribution. Built upon a deep neural network backbone, it incorporates robust feature matching and iterative optimization modules. Extensive evaluation on cross-domain benchmarks—ScanNet and MegaDepth—demonstrates significant improvements in matching accuracy. Ablation studies confirm the complementary effectiveness of all components. This work breaks the strong homogeneity assumption inherent in conventional learning-based RANSAC, enabling robust geometric estimation with strong generalization across diverse, unseen domains.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Learning with Manifolds

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Random Sample Consensus (RANSAC) is a fundamental approach for robustly estimating parametric models from noisy data. Existing learning-based RANSAC methods utilize deep learning to enhance the robustness of RANSAC against outliers. However, these approaches are trained and tested on the data generated by the same algorithms, leading to limited generalization to out-of-distribution data during inference. Therefore, in this paper, we introduce a novel diffusion-based paradigm that progressively injects noise into ground-truth data, simulating the noisy conditions for training learning-based RANSAC. To enhance data diversity, we incorporate Monte Carlo sampling into the diffusion paradigm, approximating diverse data distributions by introducing different types of randomness at multiple stages. We evaluate our approach in the context of feature matching through comprehensive experiments on the ScanNet and MegaDepth datasets. The experimental results demonstrate that our Monte Carlo diffusion mechanism significantly improves the generalization ability of learning-based RANSAC. We also develop extensive ablation studies that highlight the effectiveness of key components in our framework.

Problem

• Enhance generalization of learning-based RANSAC for out-of-distribution data.
• Introduce diffusion-based paradigm to simulate noisy training conditions.
• Incorporate Monte Carlo sampling to approximate diverse data distributions.

Innovation

• Introduces diffusion-based paradigm for RANSAC training
• Uses Monte Carlo sampling to enhance data diversity
• Improves generalization of learning-based RANSAC