Optimal transport unlocks end-to-end learning for single-molecule localization

• URL: http://arxiv.org/abs/2512.10683v1
• Date: Thu, 11 Dec 2025 14:30:16 GMT
• Title: Optimal transport unlocks end-to-end learning for single-molecule localization
• Authors: Romain Seailles, Jean-Baptiste Masson, Jean Ponce, Julien Mairal,
• Abstract summary: Single-molecule localization microscopy allows reconstructing biology-relevant structures beyond the diffraction limit.<n>Currently, efficient SMLM requires non-overlapping emitting fluorophores.<n>Recent deep-learning approaches can handle denser emissions, but they rely on variants of non-maximum suppression.<n>We propose an iterative neural network that integrates knowledge of the microscope's optical system inside our model.
• Score: 26.867838368484268
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Single-molecule localization microscopy (SMLM) allows reconstructing biology-relevant structures beyond the diffraction limit by detecting and localizing individual fluorophores -- fluorescent molecules stained onto the observed specimen -- over time to reconstruct super-resolved images. Currently, efficient SMLM requires non-overlapping emitting fluorophores, leading to long acquisition times that hinders live-cell imaging. Recent deep-learning approaches can handle denser emissions, but they rely on variants of non-maximum suppression (NMS) layers, which are unfortunately non-differentiable and may discard true positives with their local fusion strategy. In this presentation, we reformulate the SMLM training objective as a set-matching problem, deriving an optimal-transport loss that eliminates the need for NMS during inference and enables end-to-end training. Additionally, we propose an iterative neural network that integrates knowledge of the microscope's optical system inside our model. Experiments on synthetic benchmarks and real biological data show that both our new loss function and architecture surpass the state of the art at moderate and high emitter densities. Code is available at https://github.com/RSLLES/SHOT.

Related papers

• $μ$NeuFMT: Optical-Property-Adaptive Fluorescence Molecular Tomography via Implicit Neural Representation [17.533721844205687]
  Fluorescence Molecular Tomography (FMT) is a promising technique for non-invasive 3D visualization of fluorescent probes.<n>We propose $mu$NeuFMT, a self-supervised FMT reconstruction framework that integrates implicit neural-based scene representation with explicit physical modeling of photon propagation.<n>Our work establishes a new paradigm for robust and accurate FMT reconstruction, paving the way for more reliable molecular imaging in complex clinically related scenarios.
  arXiv  Detail & Related papers  (2025-11-06T16:28:30Z)
• Combining GCN Structural Learning with LLM Chemical Knowledge for Enhanced Virtual Screening [0.562479170374811]
  We propose a hybrid architecture that combines localized structural learning with global chemical knowledge.<n>We found that concatenating the LLM embeddings after each GCN layer-rather than only at the final layer-significantly improves performance.<n>The resulting model achieves superior results, with an F1-score of (88.8%), outperforming standalone GCN (87.9%), XGBoost (85.5%), and SVM (85.4%) baselines.
  arXiv  Detail & Related papers  (2025-04-24T12:38:03Z)
• CAF-YOLO: A Robust Framework for Multi-Scale Lesion Detection in Biomedical Imagery [0.0682074616451595]
  CAF-YOLO is a nimble yet robust method for medical object detection that leverages the strengths of convolutional neural networks (CNNs) and transformers. ACFM module enhances the modeling of both global and local features, enabling the capture of long-term feature dependencies. MSNN improves multi-scale information aggregation by extracting features across diverse scales.
  arXiv  Detail & Related papers  (2024-08-04T01:44:44Z)
• Solving Energy-Independent Density for CT Metal Artifact Reduction via Neural Representation [46.57879724994237]
  Reconstructing CT images from metal-corrupted measurements becomes a challenging nonlinear inverse problem.<n>Existing state-of-the-art (SOTA) metal artifact reduction (MAR) algorithms rely on supervised learning with numerous paired CT samples.<n>In this work, we propose Density neural representation (Diner), a novel unsupervised MAR method.
  arXiv  Detail & Related papers  (2024-05-11T16:30:39Z)
• Self-STORM: Deep Unrolled Self-Supervised Learning for Super-Resolution Microscopy [55.2480439325792]
  We introduce deep unrolled self-supervised learning, which alleviates the need for such data by training a sequence-specific, model-based autoencoder. Our proposed method exceeds the performance of its supervised counterparts.
  arXiv  Detail & Related papers  (2024-03-25T17:40:32Z)
• Evaluation of Activated Sludge Settling Characteristics from Microscopy Images with Deep Convolutional Neural Networks and Transfer Learning [7.636901972162706]
  This study presents an innovative computer vision-based approach to assess activated sludge-settling characteristics. Implementing the transfer learning of deep convolutional neural network (CNN) models, this approach aims to overcome the limitations of existing quantitative image analysis techniques. Various CNN architectures, including Inception v3, ResNet18, ResNet152, ConvNeXt-nano, and ConvNeXt-S, were tested to evaluate their performance in predicting sludge settling characteristics.
  arXiv  Detail & Related papers  (2024-02-14T18:13:37Z)
• Physics-Inspired Degradation Models for Hyperspectral Image Fusion [61.743696362028246]
  Most fusion methods solely focus on the fusion algorithm itself and overlook the degradation models. We propose physics-inspired degradation models (PIDM) to model the degradation of LR-HSI and HR-MSI. Our proposed PIDM can boost the fusion performance of existing fusion methods in practical scenarios.
  arXiv  Detail & Related papers  (2024-02-04T09:07:28Z)
• Learning Multiscale Consistency for Self-supervised Electron Microscopy Instance Segmentation [48.267001230607306]
  We propose a pretraining framework that enhances multiscale consistency in EM volumes. Our approach leverages a Siamese network architecture, integrating strong and weak data augmentations. It effectively captures voxel and feature consistency, showing promise for learning transferable representations for EM analysis.
  arXiv  Detail & Related papers  (2023-08-19T05:49:13Z)
• Fluctuation-based deconvolution in fluorescence microscopy using plug-and-play denoisers [2.236663830879273]
  spatial resolution of images of living samples obtained by fluorescence microscopes is physically limited due to the diffraction of visible light. Several deconvolution and super-resolution techniques have been proposed to overcome this limitation.
  arXiv  Detail & Related papers  (2023-03-20T15:43:52Z)
• An unsupervised deep learning algorithm for single-site reconstruction in quantum gas microscopes [47.187609203210705]
  In quantum gas microscopy experiments, reconstructing the site-resolved lattice occupation with high fidelity is essential for the accurate extraction of physical observables. Here, we present a novel algorithm based on deep convolutional neural networks to reconstruct the site-resolved lattice occupation with high fidelity.
  arXiv  Detail & Related papers  (2022-12-22T18:57:27Z)
• A parameter refinement method for Ptychography based on Deep Learning concepts [55.41644538483948]
  coarse parametrisation in propagation distance, position errors and partial coherence frequently menaces the experiment viability. A modern Deep Learning framework is used to correct autonomously the setup incoherences, thus improving the quality of a ptychography reconstruction. We tested our system on both synthetic datasets and also on real data acquired at the TwinMic beamline of the Elettra synchrotron facility.
  arXiv  Detail & Related papers  (2021-05-18T10:15:17Z)
• Towards an Automatic Analysis of CHO-K1 Suspension Growth in Microfluidic Single-cell Cultivation [63.94623495501023]
  We propose a novel Machine Learning architecture, which allows us to infuse a neural deep network with human-powered abstraction on the level of data. Specifically, we train a generative model simultaneously on natural and synthetic data, so that it learns a shared representation, from which a target variable, such as the cell count, can be reliably estimated.
  arXiv  Detail & Related papers  (2020-10-20T08:36:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.