ResMatching: Noise-Resilient Computational Super-Resolution via Guided Conditional Flow Matching

2025-10-30

AI Summary

Computational super-resolution (CSR) for fluorescence microscopy is a classic ill-posed inverse problem, where performance is fundamentally limited by the capacity to model priors for high-frequency information recovery—especially under low signal-to-noise ratios (SNR). To address this, we propose ResMatching, a novel method that implicitly models the image posterior distribution via guided conditional flow matching, enabling noise-robust super-resolution reconstruction. Crucially, ResMatching simultaneously produces pixel-wise uncertainty estimates, quantifying the confidence of reconstructed outputs. Evaluated on the BioSR benchmark, ResMatching achieves state-of-the-art trade-offs between fidelity and perceptual quality across four distinct biological structures, consistently outperforming seven established baselines. Its advantage is particularly pronounced in high-noise regimes, demonstrating superior robustness and reliability.

Technology Category

• Computer Vision: Medical and Biological Imaging
• Machine Learning: Calibration & Uncertainty Quantification
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Computational Super-Resolution (CSR) in fluorescence microscopy has, despite being an ill-posed problem, a long history. At its very core, CSR is about finding a prior that can be used to extrapolate frequencies in a micrograph that have never been imaged by the image-generating microscope. It stands to reason that, with the advent of better data-driven machine learning techniques, stronger prior can be learned and hence CSR can lead to better results. Here, we present ResMatching, a novel CSR method that uses guided conditional flow matching to learn such improved data-priors. We evaluate ResMatching on 4 diverse biological structures from the BioSR dataset and compare its results against 7 baselines. ResMatching consistently achieves competitive results, demonstrating in all cases the best trade-off between data fidelity and perceptual realism. We observe that CSR using ResMatching is particularly effective in cases where a strong prior is hard to learn, e.g. when the given low-resolution images contain a lot of noise. Additionally, we show that ResMatching can be used to sample from an implicitly learned posterior distribution and that this distribution is calibrated for all tested use-cases, enabling our method to deliver a pixel-wise data-uncertainty term that can guide future users to reject uncertain predictions.

Problem

• Develop noise-resilient super-resolution for fluorescence microscopy
• Learn improved data priors using guided conditional flow matching
• Provide pixel-wise uncertainty estimates for prediction reliability

Innovation

• Guided conditional flow matching for super-resolution
• Noise-resilient prior learning from low-resolution images
• Pixel-wise uncertainty estimation for prediction reliability