Toward_spatial_unbiased-Tensorflow

KO | EN

gitlite — search

#javascript #python #react #typescript #llm #golang #go #machine-learning #linux #nodejs #windows #awesome #security #ios

Toward_spatial_unbiased-Tensorflow

(★ 15)

Simple Tensorflow implementation of "Toward Spatially Unbiased Generative Models" (ICCV 2021)

Toward_spatial_unbiased-Tensorflow Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

## Spatial unbiased GANs &mdash; Simple TensorFlow Implementation [[Paper]](https://arxiv.org/abs/2108.01285)
### : Toward Spatially Unbiased Generative Models (ICCV 2021)

> **Abstract** *Recent image generation models show remarkable generation performance. However, they mirror strong location preference in datasets, which we call **spatial bias**. Therefore, generators render poor samples at unseen locations and scales. We argue that the generators rely on their implicit positional encoding to render spatial content. From our observations, the generator’s implicit positional encoding is translation-variant, making the generator spatially biased. To address this issue, we propose injecting explicit positional encoding at each scale of the generator. By learning the spatially unbiased generator, we facilitate the robust use of generators in multiple tasks, such as GAN inversion, multi-scale generation, generation of arbitrary sizes and aspect ratios. Furthermore, we show that our method can also be applied to denoising diffusion probabilistic models.*

## Requirements
* `Tensorflow >= 2.x`

## Usage
```
├── dataset
   └── YOUR_DATASET_NAME
       ├── 000001.jpg 
       ├── 000002.png
       └── ...
```

### Train
```
> python main.py --dataset FFHQ --phase train --img_size 256 --batch_size 4 --n_total_image 6400
```

### Generate Video
```
> python generate_video.py
```

## Results
* **FID: 3.81 (6.4M images(200k iterations), 8GPU, each 4 batch size)**
  * FID reported in the paper: **6.75**
### Video
<div align="center">
  <img src="./assets/sample_2.gif">
</div>

### Uncuratd
<div align="center">
  <img src="./assets/uncurated.png">
</div>

### Style mixing
* It's worse than stylegan2.
<div align="center">
  <img src="./assets/style_mixing.png">
</div>

### Truncation trick
<div align="center">
  <img src="./assets/truncation_trick.png">
</div>

## Reference
* [Official Pytorch](https://github.com/jychoi118/toward_spatial_unbiased)
* [StyleGAN2-Tensorflow](https://github.com/moono/stylegan2-tf-2.x)

## Author
[Junho Kim](http://bit.ly/jhkim_resume)

Toward_spatial_unbiased-Tensorflow

Toward_spatial_unbiased-Tensorflow Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Toward_spatial_unbiased-Tensorflow

Toward_spatial_unbiased-Tensorflow Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README