search

fdjingyuan / Deep-Fashion-Analysis-ECCV2018

Codes of ECCV 2018 workshop paper "Deep Fashion Analysis with Feature Map Upsampling and Landmark-driven Attention"
83 stars 27 forks source link

run prediction #11

Open erixavero opened 4 years ago

erixavero commented 4 years ago

once training is finished, there is a 'models' folder and pkl file created, is this the fully trained model? and how to use it to run prediction? here is my specs Ubuntu 18.04.3 LTS Cuda compilation tools, release 9.1, V9.1.85 Python 3.7.0

johnspencer98 commented 4 years ago

Yes also confused on this

indersingh17188 commented 3 years ago

Hi,

I have tried to create a small code to run predictions but still getting stuck as the predictions for example landmarks are supposed to be converted in order to be superimposed on the original image. It is getting confusing to obtain proper prediction which can be displayed over original input image for testing purposes.

model = const.USE_NET() model = model.to(const.device) model.load_state_dict(torch.load(PATH))

where PATH is the location of .pkl file. I modified forward function of lm_networks in order to read a normal image but still unable to get proper predictions. Any update will be highly appreciated.

Thanks