/Similariy-DT

Codes for the paper "Kernelized Similarity Learning and Embedding for Dynamic Texture Synthesis" Accepted to T-SMCA, 2020

Primary LanguageMATLABMIT LicenseMIT

Kernel Similarity Learning and Embedding for DT Synthesis [Project]

Overview

Installation

  1. Clone this repo.
git clone https://github.com/shiming-chen/Similariy-DT.git
cd Similarity-DT

Generation (Running on Windows)

1. Generating gray DTs

Run Similarity_gray('DT_name','save_name')

2. Generating RGB DTs

Run Similarity_RGB('DT_name','save_name')

3. Transfering Trained Model for Generation

Run Similarity_RGB_transfer('train_DT_name','test_DT_name','save_name')

Baseline

1. Non-Neural-Network-Based Methods

We provide the codes for the compared non-neural-network-based methods, i.e., LDS, SLDS, Kernel-DT, FFT-LDS, HOSVD, KPCR.

2. Neural-Network-Based Methods

The project pages of TwoStream, STGCN, and DG are presented as follow:

  1. TwoStream: https://ryersonvisionlab.github.io/two-stream-projpage/
  2. STGCN: http://www.stat.ucla.edu/~jxie/STGConvNet/STGConvNet.html
  3. DG: http://www.stat.ucla.edu/~jxie/DynamicGenerator/DynamicGenerator.html

Contact

If you run into any problems with these codes, please submit a bug report on the Github site of the project. For another inquries please contact with me: gchenshiming@gmail.com or shimingchen@hust.edu.cn.