elvisyjlin / spatialattentiongan Goto Github PK

SaGAN PyTorch "Generative Adversarial Network with Spatial Attention for Face Attribute Editing"

License: MIT License

Python 100.00%

sagan spatial-attention face-attribute-manipulation pytorch eccv-2018

spatialattentiongan's Introduction

Spatial Attention Generative Adversarial Network

This repository contains the PyTorch implementation of the ECCV 2018 paper "Generative Adversarial Network with Spatial Attention for Face Attribute Editing" (pdf).

My results with images and attention masks on CelebA 128 (original, eyeglasses, mouth_slightly_open, no_beard, smiling)

Requirements

Python 3.5
PyTorch 1.0.0

pip3 install -r requirements.txt

The training procedure described in paper takes 5.5GB memory on a single GPU.

Datasets
- CelebA
  - Put Align&Cropped Images in ./data/celeba/*.jpg
  - Put Attributes Annotations in ./data/list_attr_celeba.txt

Pretrained models (download from http://bit.ly/sagan-results and decompress the zips to ./results)

  results
  ├── celeba_128_eyeglasses
  ├── celeba_128_mouth_slightly_open
  ├── celeba_128_no_beard
  └── celeba_128_smiling

Usage

Train a model with a target attribute

python3 train.py --experiment-name celeba_128_eyeglasses --target-attr Eyeglasses --gpu

Generate images from trained models

python3 generate.py --experiment-name celeba_128_eyeglasses --gpu

spatialattentiongan's People

Contributors

Stargazers

Watchers

spatialattentiongan's Issues

how can i use it for my use case.

My use case is I'm gonna pass a face image to the GAN model and it should return the same image with eyeglasses.
can i use your code to do it ? If yes , i need some suggestion.
Thanks .

about Align&Cropped Images

hi，your work is great, but i want to know how to get the 128*128 Align&Cropped Images?

Encounting `TypeError` when execute the training script

Can you give me an example of colab notebook to train the model? since i've tried execute a script in Google Colab, however, what i've got is that

The command to run a script (in Google Colab)

!python3 /content/SpatialAttentionGAN/train.py --experiment-name celeba_128_eyeglasses --target-attr Eyeglasses --gpu

Error encounted

Namespace(data_path='./data/celeba', attr_path='./data/list_attr_celeba.txt', target_attr='Goatee', image_size=128, batch_size=16, lr=0.0002, beta1=0.5, beta2=0.999, l1=20, l2=100, lgp=10, d_iters=3, total_kimg=1000, tick_kimg=5.0, sample_ticks=1, save_ticks=10, num_samples=64, experiment_name='celeba_128_eyeglasses2', gpu=False)
Traceback (most recent call last):
  File "/content/SpatialAttentionGAN/train.py", line 119, in <module>
    vutils.save_image(fixed_reals, join(sample_path, '{:07d}_real.jpg'.format(0)), nrow=8, padding=0, normalize=True, range=(-1., 1.))
  File "/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py", line 115, in decorate_context
    return func(*args, **kwargs)
  File "/usr/local/lib/python3.10/dist-packages/torchvision/utils.py", line 146, in save_image
    grid = make_grid(tensor, **kwargs)
  File "/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py", line 115, in decorate_context
    return func(*args, **kwargs)
TypeError: make_grid() got an unexpected keyword argument 'range'

why the result of color of glass is light

hi， i have other question, that is why the result of color of glass is light

CUDA problem in generate.py

Thanks for your excellent work, but I got similar problem as #1.

Traceback (most recent call last):
  File "generate.py", line 80, in <module>
    samples, masks = G(reals, target_labels)
  File "/home/zhiyong/.local/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/zhiyong/temp_test/SpatialAttentionGAN/sagan.py", line 98, in forward
    y = self.AMN(x, a)
  File "/home/zhiyong/.local/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/zhiyong/temp_test/SpatialAttentionGAN/sagan.py", line 86, in forward
    x = torch.cat((x, a), dim=1)
RuntimeError: Expected object of backend CUDA but got backend CPU for sequence element 1 in sequence argument at position #1 'tensors'

And I fixed it by modify the code in line76, generate.py.

reals, labels = reals.to(device), labels.type_as(reals).to(device)

Maybe you can update the generate.py to fix it.

Hi!
When I run train.py, it occured:
Traceback (most recent call last):
File "train.py", line 148, in
fakes, _ = G(reals, target_labels)
File "/home/rain/software/anaconda3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 489, in call
result = self.forward(*input, **kwargs)
File "/home/rain/nlp/etrump/txt2img/SAGAN_celeba/sagan.py", line 95, in forward
y = self.AMN(x, a)
File "/home/rain/software/anaconda3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 489, in call
result = self.forward(*input, **kwargs)
File "/home/rain/nlp/etrump/txt2img/SAGAN_celeba/sagan.py", line 83, in forward
x = torch.cat((x, a), dim=1)
RuntimeError: Expected object of backend CUDA but got backend CPU for sequence element 1 in sequence argument at position #1 'tensors'

Can you help me with it?

Thank you so much!