Precomputed Real-Time Texture Synthesis with Markovian Generative Adversarial Networks

Related tags

Deep LearningMGANs
Overview

MGANs

Training & Testing code (torch), pre-trained models and supplementary materials for "Precomputed Real-Time Texture Synthesis with Markovian Generative Adversarial Networks".

See this video for a quick explaination for our method and results.

Setup

As building Torch with the latest CUDA is a troublesome work, we recommend following the following steps to people who want to reproduce the results: It has been tested on Ubuntu with CUDA 10.

Step One: Install CUDA 10 and CUDNN 7.6.2

If you have a fresh Ubuntu, we recommend Lambda Stack which helps you install the latest drivers, libraries, and frameworks for deep learning. Otherwise, you can install the CUDA toolkit and CUDNN from these links:

Step Two: Install Torch

git clone https://github.com/nagadomi/distro.git ~/torch --recursive
cd ~/torch
./install-deps
./clean.sh
./update.sh

. ~/torch/install/bin/torch-activate
sudo apt-get install libprotobuf-dev protobuf-compiler
luarocks install loadcaffe

Demo

cd code
th demo_MGAN.lua

Training

Simply cd into folder "code/" and run the training script.

th train.lua

The current script is an example of training a network from 100 ImageNet photos and a single painting from Van Gogh. The input data are organized in the following way:

  • "Dataset/VG_Alpilles_ImageNet100/ContentInitial": 5 training ImageNet photos to initialize the discriminator.
  • "Dataset/VG_Alpilles_ImageNet100/ContentTrain": 100 training ImageNet photos.
  • "Dataset/VG_Alpilles_ImageNet100/ContentTest": 10 testing ImageNet photos (for later inspection).
  • "Dataset/VG_Alpilles_ImageNet100/Style": Van Gogh's painting.

The training process has three main steps:

  • Use MDAN to generate training images (MDAN_wrapper.lua).
  • Data Augmentation (AG_wrapper.lua).
  • Train MGAN (MDAN_wrapper.lua).

Testing

The testing process has two steps:

  • Step 1: call "th release_MGAN.lua" to concatenate the VGG encoder with the generator.
  • Step 2: call "th demo_MGAN.lua" to test the network with new photos.

Display

You can use the browser based display package to display the training process for both MDANs and MGANs.

  • Install: luarocks install https://raw.githubusercontent.com/szym/display/master/display-scm-0.rockspec
  • Call: th -ldisplay.start
  • See results at this URL: http://localhost:8000

Example

We chose Van Gogh's "Olive Trees with the Alpilles in the Background" as the reference texture.

We then transfer 100 ImageNet photos into the same style with the proposed MDANs method. MDANs take an iterative deconvolutional approach, which is similar to "A Neural Algorithm of Artistic Style" by Leon A. Gatys et al. and our previous work "CNNMRF". Differently, it uses adversarial training instead of gaussian statistics ("A Neural Algorithm of Artistic Style) or nearest neighbour search "CNNMRF". Here are some transferred results from MDANs:

The results look nice, so we know adversarial training is able to produce results that are comparable to previous methods. In other experiments we observed that gaussian statistics work remarkable well for painterly textures, but can sometimes be too flexible for photorealistic textures; nearest-neighbor search preserve photorealistic details but can be too rigid for deformable textures. In some sense MDANs offers a relatively more balanced choice with advaserial training. See our paper for more discussoins.

Like previous deconvolutional methods, MDANs is VERY slow. A Nvidia Titan X takes about one minute to transfer a photo of 384 squared. To make it faster, we replace the deconvolutional process by a feed-forward network (MGANs). The feed-forward network takes long time to train (45 minutes for this example on a Titan X), but offers significant speed up in testing time. Here are some results from MGANs:

It is our expectation that MGANs will trade quality for speed. The question is: how much? Here are some comparisons between the result of MDANs and MGANs:

In general MDANs (middle) give more stylished results, and does a much better job at homegenous background areas (the last two cases). But sometimes MGANs (right) is able to produce comparable results (the first two).

And MGANs run at least two orders of magnitudes faster.

Final remark

There are concurrent works that try to make deep texture synthesis faster. For example, Ulyanov et al. and Johnson et al. also achieved significant speed up and very nice results with a feed-forward architecture. Both of these two methods used the gaussian statsitsics constraint proposed by Gatys et al.. We believe our method is a good complementary: by changing the gaussian statistics constraint to discrimnative networks trained with Markovian patches, it is possible to model more complex texture manifolds (see discussion in our paper).

Last, here are some prelimiary results of training a MGANs for photorealistic synthesis. It learns from 200k face images from CelebA. The network then transfers VGG_19 encoding (layer ReLU5_1) of new face images (left) into something interesting (right). The synthesized faces have the same poses/layouts as the input faces, but look like different persons :-)

Acknowledgement

Code for the paper: On Pathologies in KL-Regularized Reinforcement Learning from Expert Demonstrations

Non-Parametric Prior Actor-Critic (N-PPAC) This repository contains the code for On Pathologies in KL-Regularized Reinforcement Learning from Expert D

Cong Lu 5 May 13, 2022
An interactive DNN Model deployed on web that predicts the chance of heart failure for a patient with an accuracy of 98%

Heart Failure Predictor About A Web UI deployed Dense Neural Network Model Made using Tensorflow that predicts whether the patient is healthy or has c

Adit Ahmedabadi 0 Jan 09, 2022
Source code for Transformer-based Multi-task Learning for Disaster Tweet Categorisation (UCD's participation in TREC-IS 2020A, 2020B and 2021A).

Source code for "UCD participation in TREC-IS 2020A, 2020B and 2021A". *** update at: 2021/05/25 This repo so far relates to the following work: Trans

Congcong Wang 4 Oct 19, 2021
Multi-Agent Reinforcement Learning (MARL) method to learn scalable control polices for multi-agent target tracking.

scalableMARL Scalable Reinforcement Learning Policies for Multi-Agent Control CD. Hsu, H. Jeong, GJ. Pappas, P. Chaudhari. "Scalable Reinforcement Lea

Christopher Hsu 17 Nov 17, 2022
Python framework for Stochastic Differential Equations modeling

SDElearn: a Python package for SDE modeling This package implements functionalities for working with Stochastic Differential Equations models (SDEs fo

4 May 10, 2022
Res2Net for Instance segmentation and Object detection using MaskRCNN

Res2Net for Instance segmentation and Object detection using MaskRCNN Since the MaskRCNN-benchmark of facebook is deprecated, we suggest to use our mm

Res2Net Applications 55 Oct 30, 2022
Official implementation of the paper 'Efficient and Degradation-Adaptive Network for Real-World Image Super-Resolution'

DASR Paper Efficient and Degradation-Adaptive Network for Real-World Image Super-Resolution Jie Liang, Hui Zeng, and Lei Zhang. In arxiv preprint. Abs

81 Dec 28, 2022
A curated list of long-tailed recognition resources.

Awesome Long-tailed Recognition A curated list of long-tailed recognition and related resources. Please feel free to pull requests or open an issue to

Zhiwei ZHANG 542 Jan 01, 2023
Using VideoBERT to tackle video prediction

VideoBERT This repo reproduces the results of VideoBERT (https://arxiv.org/pdf/1904.01766.pdf). Inspiration was taken from https://github.com/MDSKUL/M

75 Dec 14, 2022
Official Implementation for Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation

Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation We present a generic image-to-image translation framework, pixel2style2pixel (pSp

2.8k Dec 30, 2022
A Python package for time series augmentation

tsaug tsaug is a Python package for time series augmentation. It offers a set of augmentation methods for time series, as well as a simple API to conn

Arundo Analytics 278 Jan 01, 2023
PyTorch implementation of PNASNet-5 on ImageNet

PNASNet.pytorch PyTorch implementation of PNASNet-5. Specifically, PyTorch code from this repository is adapted to completely match both my implemetat

Chenxi Liu 314 Nov 25, 2022
This framework implements the data poisoning method found in the paper Adversarial Examples Make Strong Poisons

Adversarial poison generation and evaluation. This framework implements the data poisoning method found in the paper Adversarial Examples Make Strong

31 Nov 01, 2022
PyTorch implementation of Deep HDR Imaging via A Non-Local Network (TIP 2020).

NHDRRNet-PyTorch This is the PyTorch implementation of Deep HDR Imaging via A Non-Local Network (TIP 2020). 0. Differences between Original Paper and

Yutong Zhang 1 Mar 01, 2022
FedJAX is a library for developing custom Federated Learning (FL) algorithms in JAX.

FedJAX: Federated learning with JAX What is FedJAX? FedJAX is a library for developing custom Federated Learning (FL) algorithms in JAX. FedJAX priori

Google 208 Dec 14, 2022
A modular active learning framework for Python

Modular Active Learning framework for Python3 Page contents Introduction Active learning from bird's-eye view modAL in action From zero to one in a fe

modAL 1.9k Dec 31, 2022
A minimalist tool to display a network graph.

A tool to get a minimalist view of any architecture This tool has only be tested with the models included in this repo. Therefore, I can't guarantee t

Thibault Castells 1 Feb 11, 2022
Measures input lag without dedicated hardware, performing motion detection on recorded or live video

What is InputLagTimer? This tool can measure input lag by analyzing a video where both the game controller and the game screen can be seen on a webcam

Bruno Gonzalez 4 Aug 18, 2022
Code, final versions, and information on the Sparkfun Graphical Datasheets

Graphical Datasheets Code, final versions, and information on the SparkFun Graphical Datasheets. Generated Cells After Running Script Example Complete

SparkFun Electronics 102 Jan 05, 2023
Self-Attention Between Datapoints: Going Beyond Individual Input-Output Pairs in Deep Learning

We challenge a common assumption underlying most supervised deep learning: that a model makes a prediction depending only on its parameters and the features of a single input. To this end, we introdu

OATML 360 Dec 28, 2022