# memgan-pytorch

**Repository Path**: swu-wzt/memgan-pytorch

## Basic Information

- **Project Name**: memgan-pytorch
- **Description**: Release code for "MemoryGAN: GAN Generator as Heterogeneous Memory for compositional Image Synthesis "
- **Primary Language**: Unknown
- **License**: MulanPSL-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2023-06-30
- **Last Updated**: 2023-07-03

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# memgan-pytorch

#### Introduction

This repo is for our work "MemoryGAN: GAN Generator as Heterogeneous Memory for Compositional Image Synthesis"


> The Generative Adversarial Network (GAN) has recently experienced great progress in
> compositional image synthesis. Unfortunately, the models proposed in the literature usually require a
> set of pre-defined local generators and use a separate generator to model each part object. This makes
> the model inflexible and also limits its scalability. Inspired by humans’ structured memory system,
> we propose MemoryGAN to eliminate these disadvantages. MemoryGAN uses a single generator as
> a shared memory to hold the heterogeneous information of the parts, and it uses a recurrent neural
> network to model the dependency between the parts and provide the query code for the memory. The
> shared memory structure and the query and feedback mechanism make MemoryGAN flexible and
> scalable. Our experiment shows that although MemoryGAN only uses a single generator for all the
> parts, it achieves comparable performance with the state-of-the-art, which uses multiple generators,
> in terms of synthesized image quality, compositional ability and disentanglement property. We
> believe that our result of using the generator of the GAN as a memory model will inspire future work
> of both bio-friendly models and memory-augmented models.


#### Installation

This repo is developed using pytorch of version 1.13, cudatoolkit vision 11.6 and nvdia driver 510 on ubuntu 20.04. Our experiments are performed using single RTX3090.

#### Usage

Please check the `train.py` file for the configuration options. To train the model, basically you only need to run:

`python train.py --YOUR-CONFIGURATION`

More details and pretrained models will be release soon ...