Fig. 2: Multi-view variational reconstructions.
From: Towards end-to-end structure determination from x-ray diffraction data using deep learning
We display CMED reconstructions of previously unseen crystals at multiple viewing angles. Given chemical composition and powder XRD as input, we generate a distribution over latent codes, which we then sample from to obtain multiple crystal reconstructions. This allows us to obtain better guesses, if the first prediction is sub-optimal.
