Software for Aligned Cluster Analysis

DOWNLOAD

Introduction

This page contains software and instructions for aligned cluster analysis (ACA)[2] and hierarchical aligned cluster analysis (HACA)[1].

Installation

1 2   3 4 5
>> cd aca >> ls make.m addPath.m demoMocap.m demoToy.m demoToyH.m readme data lib src >> make >> addPath >> demoMocap

Package Content

The package of aca.zip contains following folders and files:
data: This folder contains a subset of CMU Motion Capture dataset.
src: This folder contains the main implementation of ACA and HACA.
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 clustering of a synthetic sequence by ACA.
demoToyH.m: Time clustering of a synthetic sequence by HACA.
demoMocap.m: Time clustering of motion capture sequence by ACA and HACA.
By using this function, you can obtain results similar to those shown here.

FAQs

Which functions are implemented in C++?

Most of the functions have been written and documented in Matlab format. We additionally provide C++ implementations of some dynamic programming routines which involve many loops and are notoriously slow in Matlab. For each C++ code, we provide its corresponding Matlab version. For instance, you can use acaFordSlow.m instead of acaFord.cpp. They have the same interface in both input and output. The C++ code is faster to obtain result while the Matlab version is easier to understand and debug.

How to locally constrain the time warping step in DTAK?

In the code, we provided another version of DTAK algorithm in which the dynamic programming can be locally constrained. Although we didn't use this feature in our paper, we found it is useful to obtain a robust alignment and to speedup the algorithm. Please refer to this paper[3] for more details on constraints in dynamic time warping (DTW).

Change Log

References

Copyright

This software is free for use in research projects. If you publish results obtained using this software, please use this citation:

@article{ZhouDH13,     author = {Feng Zhou and Fernando {De la Torre} and Jessica K. Hodgins},     title = {Hierarchical Aligned Cluster Analysis for Temporal Clustering of Human Motion},     journal = {IEEE Transactions Pattern Analysis and Machine Intelligence (PAMI)},     year = {2013},     volume = {35},     number = {3},     pages = {582-596}, }

Contributing back bug fixes and improvements is polite and encouraged. If you have any question, feel free to contact Feng Zhou.