Explore high-quality datasets for your AI and machine learning projects.
This dataset is primarily intended for image-to-image tasks and involves lensless and computational imaging technologies. It contains two main feature types, *lensless* and *lensed*, both of which are image data. The dataset is split into a training set of 24,000 samples and a test set of 999 samples. The download size is 5,873,531,153 bytes and the stored size is 5,831,439,790 bytes. It is a replica of the original DiffuserCam MirFlickr dataset, provided as NPY files for training and processed with a minimum‑value subtraction when saved as viewable images.
