Fig. 7: CrystalNet system diagram.
From: Towards end-to-end structure determination from x-ray diffraction data using deep learning
XRD and chemical input are passed through encoders, which output latent feature distributions. Sampling from those latent distributions, we fuse the features with the positional (x, y, z) query, and pass the fused information through the charge density regressor to get our final charge density prediction.
