Benchmarking Deep Inverse Models over time, and the Neural-Adjoint method

Neural Information Processing Systems 

We find that the addition of a simple novel loss term - which we term the boundary loss - dramatically improves the NA's