Question on pre-trained models

Div99 / W-Stereo-Disp

(NeurIPS 2020 Spotlight) Wasserstein Distances for Stereo Disparity Estimation

https://div99.github.io/W-Stereo-Disp/

MIT License

101 stars 15 forks source link

Question on pre-trained models #6

Closed Nicholasli1995 closed 3 years ago

Nicholasli1995 commented 3 years ago

Hi, thank you for releasing the pre-trained depth models. Could you please advise if there is a script to re-produce CDN-DSGN in the current repo? Or are there instructions to do so?

mgladkova commented 3 years ago

Also interested!

Div99 commented 3 years ago

Hi, it is very straightforward to modify DSGN similar to how we modify vanilla PSMNet to our new version.

Essentially you need to directly return the cost volume output and train it with the W_loss.

Div99 commented 3 years ago

I will try to release the CDN-DSGN code when I get time