Harold Phelippeau

TomoSegNet improves membrane segmentation in cryo-electron tomography by using synthetic data generated with a biophysical simulator, outperforming existing methods in recovering distorted membranes.

This paper introduces a deep learning model for orientation estimation of macromolecules in cryo-ET images using synthetic data and shows promising results.

This paper introduces synthetic datasets for cryo-electron tomography to train deep learning algorithms by simulating cellular structures and geometries.