This is a data set that contains 300 labelled HeLa cell images, indicating the five different classes: nuclear envelope, nucleus, other cells, background and cell. The ground truth is based on a data set previously released in Zenodo (https://doi.org/10.5281/zenodo.3874949), which contained four classes, nuclear envelope, nucleus, cells, background. For this data set the class cell was divided to contain other cells and the cell that corresponds to the nucleus. Each image corresponds to a 3D stack of images acquired with Electron Microscopy and are released as tif images and Matlab files. Details of the imaging, preparation and analysis of the Electron Microscopy images and have been published in: Cefa Karabağ, Martin L. Jones, Christopher J. Peddie, Anne E. Weston, Lucy M. Collinson, Constantino Carlos Reyes-Aldasoro. Segmentation and Modelling of the Nuclear Envelope of HeLa Cells Imaged with Serial Block Face Scanning Electron Microscopy. J. Imaging 2019, 5(9), 75; https://doi.org/10.3390/jimaging5090075 Karabağ C, Jones ML, Peddie CJ, Weston AE, Collinson LM, Reyes-Aldasoro, CC., Semantic segmentation of HeLa cells: An objective comparison between one traditional algorithm and four deep-learning architectures (2020). PLOS ONE 15(10): e0230605. https://doi.org/10.1371/journal.pone.0230605 The Electron Microscope images are freely available through EMPIAR: http://dx.doi.org/10.6019/EMPIAR-10094 EMPIAR.