search

cc-ai / kdb

AI can help Visualizing the Impacts of Climate Change. This is an open forum to share our work
https://mila.quebec/en/ai-society/visualizing-climate-change/
Apache License 2.0
19 stars 3 forks source link

Hardcoded domain adaptation classifier in MUNIT/SPADE #125

Open 51N84D opened 4 years ago

51N84D commented 4 years ago

The domain adaptation classifier in the MUNIT/SPADE codebases (in "utils.py") doesn't work with arbitrary latent vectors.

For example, changing the number of downsampling layers breaks the code

51N84D commented 4 years ago

Temporary solution is modifying the architecture when changing the number of downsampling layers

sashavor commented 4 years ago

Which entails...?

On Fri., Apr. 10, 2020, 5:18 p.m. Sunand Raghupathi, < notifications@github.com> wrote:

The domain adaptation classifier in the MUNIT/SPADE codebases (in "utils.py") doesn't work with arbitrary latent vectors.

For example, changing the number of downsampling layers breaks the code

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/cc-ai/kdb/issues/125, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADMMIITNWGVRLIEBP2BXIWTRL6EKNANCNFSM4MFWW35A .

sashavor commented 4 years ago

That is a lot of possible combinations though, right?

On Fri., Apr. 10, 2020, 5:19 p.m. Sunand Raghupathi, < notifications@github.com> wrote:

Temporary solution is modifying the architecture when changing the number of downsampling layers

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/cc-ai/kdb/issues/125#issuecomment-612222623, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADMMIISPATROXM46VZRCUU3RL6EN7ANCNFSM4MFWW35A .

vict0rsch commented 4 years ago

Some libraries forward 1 sample through networks to make this kind of decision. So one solution would be to give a sample to the networks creation procedure and infer the proper shapes given the sample and the config