English | 简体中文

Mnist CGAN

Background

This is an implementation of Conditional Generative Adversarial Nets based on the MINST dataset.

The MNIST data set comes from the National Institute of Standards and Technology, National Institute of Standards and Technology (NIST). The training set (training set) consists of digits handwritten by 250 different people, of which 50% are high school students and 50% are from the population Census Bureau (the Census Bureau) staff. The test set (test set) is also the same proportion of handwritten digit data.

API

The interface class file aigcmn.py implements the interface class AiGcMn. The interface class AiGcMn provides an interface function generate. The parameter of this function is an integer $n$-dimensional tensor ( $n$ is the size of the batch, each integer is in the range of 0~9, representing the number to be generated), and the output is $n12828$ tensor ( $n$ is the size of the batch, each $128*28$ tensor represents a randomly generated digital image).

Usage

This is a pytorch project so it could be used as follow:

pip install -r requirements.txt
python3 aigcmn.py

Then input() of class AiGnMn will be called to transfer input integer (0 ~ 9) to a tensor. Afterwards, method generate() will take in the tensor and returns $n128*28$ tensor. Both the output tensor text and image could be found in /result.

Contributors

This project exists thanks to all the people who contribute: Fannyzzzz, Skyuan07, chatterboxthur, noiho, young1881

License

MIT © Wortox Young