search

zhaoyuzhi / ChildPredictor

ChildPredictor: A Child Face Prediction Framework with Disentangled Learning. IEEE TMM, 2022
MIT License
53 stars 7 forks source link

increase output size #5

Open hiogo9 opened 1 year ago

hiogo9 commented 1 year ago

Hi, thanks for great model! Suffered a little with dependencies, but launched it on 11 cuda and 2.0.1 pytorch. All works great, but I have a question: is it possible to increase the size of the output image, or it was set during training and cannot be changed?

zhaoyuzhi commented 1 year ago

Yes, the output resolution is fixed during training and at inference.

Maybe an upsampler can be used like StableDiffusion and BSRGAN for post-processing ChildPredictor's output.