Load data set(s) from the Large-Scale Understanding Challenge (LSUN) (Yu et al., 2015). It consists of images in 10 different categories, each with a variable amount of training images (~100,000 to ~3 million) and 300 validation images. There is a fixed set of 10,000 test images.


  • path: str. Path to directory which either stores file or otherwise file will be downloaded and extracted there. Filenames are category + _train_lmdb/, category + _test_lmdb/, category + _val_lmdb/.
  • category: str or list of str. Category from LSUN. Can be one of "bedroom", "bridge", "church_outdoor", "classroom", "conference_room", "dining_room", "kitchen", "living_room", "restaurant", "tower".


str. It is a message advising to load data manually.

Yu, F., Seff, A., Zhang, Y., Song, S., Funkhouser, T., & Xiao, J. (2015). LSUN: Construction of a large-scale image dataset using deep learning with humans in the loop. ArXiv Preprint ArXiv:1506.03365.