gtw

gtw copied to clipboard

Introduction

This page contains software and instructions for generalized time warping (GTW) [1]. In addition, we provide implementations of the dynamic time warping (DTW) [2], derivative dynamic time warping (DDTW) [3], iterative motion warping (IMW) [4] as baselines. in order to align more than two sequences, we extended DTW, DDTW and IMW to pDTW, pDDTW and pIMW resepctively by adopting the framework of Procrustes analysis [5]. Most of the functions have been written and documented in Matlab format. In order to achieve efficiency, we provided C++ implementations for some functions.

[1] F. Zhou and F. De la Torre, "Generalized Time Warping for Multi-modal Alignment of Human Motion," in CVPR, 2012.

[2] L. Rabiner and B. Juang, "Fundamentals of speech recognition," Prentice Hall, 1993.

[3] E. J. Keogh and M. J. Pazzani, "Derivative Dynamic Time Warping," in SDM, 2001.

[4] E. Hsu and K. Pulli and J. Popovic, "Style Translation for Human Motion," ACM Trans. Graph., vol. 24, pp. 1082-1089, 2005.

[5] I. L. Dryden and K. V. Mardia, "Statistical shape analysis", Wiley, 1998.

Installation 1. unzip gtw.zip to your folder; 2. Run make.m to compile all C++ files; 3. Run addPath.m to add sub-directories into the path of Matlab. 4. Run demoToy.m

Instructions The package of gtw.zip contains following files: ./src: This folder contains the main implmentation of GTW and other baseline methods. ./lib: This folder contains some necessary library functions. ./make.m: Matlab makefile for C++ code. ./addPath.m: Adds the sub-directories into the path of Matlab. ./demoToy.m: Time alignment of three synthetic sequences by GTW and other baseline method. Same function used for replicating the first experiment in the paper [1].

Copyright This software is free for use in research projects. If you publish results obtained using this software, please use this citation. @inproceedings{ZhouD12, author = {Feng Zhou and Fernando {De la Torre}}, title = {Generalized Time Warping for Alignment of Human Motion}, booktitle = {CVPR}, year = {2012}, } If you have any question, please feel free to contact Feng Zhou ([email protected]).