Code for ICLR2018 paper: Improving GAN Training via Binarized Representation Entropy (BRE) Regularization - Y. Cao · W Ding · Y.C. Lui · R. Huang

Last update: Oct 12, 2020

Related tags

Deep Learning bre-gan

Overview

code for "Improving GAN Training via Binarized Representation Entropy (BRE) Regularization" (ICLR2018 paper)

paper: https://arxiv.org/abs/1805.03644

Getting data

For cifar10

Get the matlab version of cifar10 from https://www.cs.toronto.edu/~kriz/cifar.html run python utils/preprocess_cifar10.py [path_to_downloaded_file]

For celeba

Get the Align&Cropped version of celeba from http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html run python utils/preprocess_celeba.py [path_to_downloaded_file]

Dependencies

tensorflow (need 1.4.0 for the evaluation code to run properly) lasagne (tested with 0.2.dev1) theano (tested with 1.0.1) cuda (tested with version 8.0) cudnn (tested with 7005)

Version probably doens't matter: joblib scipy numpy PIL

Train GAN:

To run a default experiment (version of dcgan used in paper, bre turned on, cifar10 experiments)

python run.py

To reproduce results from the paper for DCGAN equal size for example:

python run.py max_iteration:40000 bre_w:1 model_func:dcgan_equal_size monitor:1

To turn off BRE for baseline:

python run.py max_iteration:40000 bre_w:0 model_func:dcgan_equal_size monitor:1

Note:

FID scoring is added, which was not used and reported at the time of publication

Code for ICLR2018 paper: Improving GAN Training via Binarized Representation Entropy (BRE) Regularization - Y. Cao · W Ding · Y.C. Lui · R. Huang

Related tags

Overview

code for "Improving GAN Training via Binarized Representation Entropy (BRE) Regularization" (ICLR2018 paper)

Getting data

For cifar10

For celeba

Dependencies

Train GAN:

Note:

Owner

General neural ODE and DAE modules for power system dynamic modeling.

RTS3D: Real-time Stereo 3D Detection from 4D Feature-Consistency Embedding Space for Autonomous Driving

Official implementation of "MetaSDF: Meta-learning Signed Distance Functions"

AdaFocus (ICCV 2021) Adaptive Focus for Efficient Video Recognition

Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis

Informal Persian Universal Dependency Treebank

[ICCV 2021] Official PyTorch implementation for Deep Relational Metric Learning.

Train CPPNs as a Generative Model, using Generative Adversarial Networks and Variational Autoencoder techniques to produce high resolution images.

Alpha-Zero - Telegram Group Manager Bot Written In Python Using Pyrogram

Code release for our paper, "SimNet: Enabling Robust Unknown Object Manipulation from Pure Synthetic Data via Stereo"

No-reference Image Quality Assessment(NIQA) Algorithms (BRISQUE, NIQE, PIQE, RankIQA, MetaIQA)

Posterior predictive distributions quantify uncertainties ignored by point estimates.

Implémentation en pyhton de l'article Depixelizing pixel art de Johannes Kopf et Dani Lischinski

Continuous Security Group Rule Change Detection & Response at scale

Logistic Bandit experiments. Official code for the paper "Jointly Efficient and Optimal Algorithms for Logistic Bandits".

code for paper "Does Unsupervised Architecture Representation Learning Help Neural Architecture Search?"

A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch

This is a Python wrapper for TA-LIB based on Cython instead of SWIG.

A hand tracking demo made with mediapipe where you can control lights with pinching your fingers and moving your hand up/down.