/Deep-Convolutional-Generative-Adversarial-Network

Tensorflow 2. This repository demonstrates how to generate images of handwritten digits (MINIST) using a Deep Convolutional Generative Adversarial Network (DCGAN). 深度卷积生成对抗网络

Primary LanguageJupyter Notebook

Deep-Convolutional-Generative-Adversarial-Network

This repository demonstrates how to generate images of handwritten digits using a Deep Convolutional Generative Adversarial Network (DCGAN).

An implementation of 2015 DCGAN Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks

Images generated during model training

File description

file name content
dataset_utils.py
train_dcgan_model.py
inference_by_dcgan_model.py
training_checkpoints Save model weight
train_epoch_produce_images The picture generated during the training of the model is generated one per training period. Used to check the effect of the model to generate a picture.
inference_produce_images Save images generated using trained models

Require

  • python 3+
  • tensorflow version 2

Code usage

Training_checkpoints already has trained model weights that can be used directly to execute 3. Model Inference.

1. Prepare Data

python dataset_utils.py

2. Train Model

python train_dcgan_model.py

3. Model Inference

python inference_by_dcgan_model.py

Image generated image effects

EPOCHS = 50