Inference - Getting non-normalized landmarks

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

Inference - Getting non-normalized landmarks #66

Open apecundo opened 1 year ago

apecundo commented 1 year ago

Hi! Thank you very much for creating this as I am also testing PIPNet.

Is there anyway to generate the landmark outputs as unnormalized? For reference, I am using the 68 pt landmark format of 300W

Thanks!

wowowowowoo commented 1 month ago

i met the same problem，do you solve it？