MicSim_FluoCurv: Synthetic microscopy images of fluorescent microtubules and their curvature maps (Ait Laydi et al., 2026 ICPR)

These datasets consist of synthetic images that mimic microscopy images of fluorescently labeled microtubules and the corresponding microtubule curvature maps, where local curvature are encoded in grey scale. These datasets support the work developed in the article by Ait Laydi et al (2026, ICPR), where they were used to train various deep learning architectures for predicting microtubule curvatures in a direct manner, without a segmenting step. Two datasets are shared: (1) The 'simple' dataset contains 1000 images where the fluorescence along each microtubule is uniform ; (2) The 'complex' dataset also contains 1000 images with the same ground truth as the 'easy' dataset, but the fluorescence along each microtubule decreases towards the extremities. As a result, exctracting the microtubule curvatures is more challenging in the 'hard' dataset, particularly at the extremities. Please cite the reference below when using the dataset: Ait Laydi et al., "MTCurv: Deep learning for direct microtubule curvature mapping in noisy fluorescence microscopy images", 2026, 28th International Conference on Pattern Recognition. This work was carried out as part of a collaborative project between Hélène Bouvrais from Institute of Genetics and Development of Rennes (CNRS, University of Rennes, France) and Yousef El Mourabit from TIAD Laboratory, FST_BM (Sultan Moulay Slimane University, Morocco).