The official PyTorch implementation of the paper "Learning by Analogy: Reliable Supervision from Transformations for Unsupervised Optical Flow Estimation".
Yes, I did. The results in Table 1 were trained on the complete Sintel train dataset.
Since OT significantly increases training time, I did not use it in most of ablations, but it can still bring performance gains(refer to Table 4). The results in Table 1 used all kinds of transformations.
You can reproduce the results just use the default parameters.
Hi, first of all great work and fantastic code! I'm trying to recreate your reported results on the Sintel Training:
Thanks!