search

ThibaultGROUEIX / AtlasNet

This repository contains the source codes for the paper "AtlasNet: A Papier-Mâché Approach to Learning 3D Surface Generation ". The network is able to synthesize a mesh (point cloud + connectivity) from a low-resolution point cloud, or from an image.
http://imagine.enpc.fr/~groueixt/atlasnet/
MIT License
677 stars 119 forks source link

Changing image path #56

Closed mschweig closed 4 years ago

mschweig commented 4 years ago

Hi, I am trying to change the image for the demo. According to readme following syntax should be working, right? python train.py --demo --demo_input_path /content/AtlasNet/doc/pictures/plane_input_demo.png --reload_model_path /content/AtlasNet/training/trained_models/atlasnet_autoencoder_25_squares/network.pth

Unfortunately, I am getting a error: 

File "train.py", line 18, in <module>
    trainer = trainer.Trainer(opt)
  File "/content/AtlasNet/training/trainer.py", line 29, in __init__
    os.mkdir(self.opt.training_media_path)
FileNotFoundError: [Errno 2] No such file or directory: 'log/02020-11-19T11:57:59.131672/training_media'

Could you please help me changing the image.

Thank you.

ThibaultGROUEIX commented 4 years ago

Hi @mschweig ,

It fails because you are missing all the network specs with --reload_model_path (it would need much more parameters). You can use dir_name instead which loads everything the model needs.

For instance, you can you : 

python train.py --demo --demo_input_path doc/pictures/visdom2.png --dir_name ./training/trained_models/atlasnet_autoencoder_25_squares/

Best regards, -Thibault