Implementing Differentiable Optimal Transport: A Case Study

About this lesson

In this video we present a case study on implementing differentiable optimal transport in PyTorch. We will look at two different methods. The first method, unrolling the forward pass optimization algorithm, is easy to implement but slow and memory intensive in the backward pass. The second method, uses implicit differentiation based from deep declarative networks. It requires more work to implement but is fast and very memory efficient. Code and examples available at https://github.com/anucvml/ddn.

Original Description