DefTruth / torchlm

💎A high level pipeline for face landmarks detection, it supports training, evaluating, exporting, inference(Python/C++) and 100+ data augmentations, can easily install via pip.
https://github.com/DefTruth/torchlm
MIT License
244 stars 25 forks source link

Will PipNet allow gradients to flow through? #64

Open mdda opened 2 years ago

mdda commented 2 years ago

I will just freeze the weights in the PipNet model itself, but I'd like to get gradients (to adjust aspects of the input image) from a loss created by the Landmarks vs some target Landmarks.

It looks like the code 'flows through' from input image to Landmark predictions - does this sound right? Or is there some strange discretisation step that would make the gradients nonsense?