4 papers
Noise-Robust Conditional Flow Matching: Generating Clean Samples from Noisy Datasets
Adrian Urbański, Gabriel della Maggiora, Artur Yakimovich
Generative models learn the statistical properties of their training data, so high-quality generation depends on clean and representative datasets. In scientific imaging, acquisiti…
Metric-Guided Synthetic Image Data Rendering for Deep Learning compatible with Agentic AI
Martina Radoynova, Samuel Pantze, Trina De +2
Deep learning computer vision for scientific applications requires collecting and annotating large datasets in a laborious, expensive and error-prone process. Synthetic data genera…
Cryo-SWAN: the Multi-Scale Wavelet-decomposition-inspired Autoencoder Network for molecular density representation of molecular volumes
Rui Li, Artsemi Yushkevich, Mikhail Kudryashev +1
Learning robust representations of 3D shapes from voxelized data is essential for advancing AI methods in biomedical imaging. However, most contemporary 3D computer vision approach…
Single-shot Star-convex Polygon-based Instance Segmentation for Spatially-correlated Biomedical Objects
Trina De, Adrian Urbanski, Artur Yakimovich
Biomedical images often contain objects known to be spatially correlated or nested due to their inherent properties, leading to semantic relations. Examples include cell nuclei bei…