AME 598 ST: Motion Capture & Analysis, Spring 2004

Course

Basics

Objectives:

Introduce the basic motion capture procedures using marker-based systems, fundamental mathematics and geometry for human pose/movement modeling, and design of vision-based (marker-less) motion capture across disciplines.

Time:

TTH 2:40−3:55pm

Venue:

Matthews Center 222

Instructor:

Professor Gang Qian

Arts, Media and Engineering Program and

Department of Electrical Engineering,

Matthews Center 240A

Arizona State University, Tempe AZ 85282-8709

Email: gang.qian@asu.edu

Tel: (480) 965-3704

Fax: (480) 965-0961

Office hrs:

TTH, 4:00pm−5:00pm (or send emails for appointments)

Recommended Readings:

Understanding Motion Capture for Computer Animation and Video Games by Alberto Menache

Assigned papers

Websites:

http://www.public.asu.edu/~gqian/AME598-ST-Spring2004.htm, course information, such as reading materials, homework, notices. Check this site often for important announcements.

Tentative

Course Contents

Marker-based Motion Capture

            Overview

            Motion capture procedures

            Effective motion capture session

            Data Cleaning

Representation and Modeling of Human movements

            Human body models,

            rotation representations

            Statistical learning (HMM/EM)

Recognition

            Statistical inference, Maximum Likelihood, Sequential Monte Carlo (SMC)

Vision-based Motion Capture

            Overview

Perspective Projection

            Edge Detection

            Matching Distances

            Tracking

            Structure-from-motion

            Motion capture from Videos using learnt movement models

            Advances in SMC for motion capture

Homework

Homework will be assigned approximately once two weeks.  Please turn in your homework before class on the due date. No late homework will be accepted.  

Exams

There will be one in-class midterm. No make-up exams will be granted, except approved in advance by myself or in the case of a legitimate unforeseen emergency. There will be no final exams, but a term project.

Grading

Homework 10%

Midterm 15%

Projects 75%

Lectures
01/20/2004 History of motion capture, Review of motion capture systems
01/22/2004 Basics on Human Kinematics (I)
01/27/2004 Basics on Human Kinematics (II)
01/29/2004 Rotation Representations, Lecture ppt (homework assigned, due on Thursday, 02/12/2004) Assigned papers
02/03/2004 Hidden Markov Models (I) (homework assigned, due on Thursday, 02/12/2004)
02/05/2004 Hidden Markov Models (II) : E-M Algorithm (homework assigned, due on Thursday, 02/19/2004, Pre-study Hidden Markov Toolkit)
02/10/2004 Sequential Monte Carlo (I)
02/12/2004

HTK, Lecture ppt

Reading assignments: HTK Manual

Chapters 1, 4, 5.10.1, 5.14, 5.15, 6, 7.1-7.6, 8.1-8.4, 11.1-11.2, 12.1, 12.2, 12.7, 12.9, 13.1-13.4, and use chapters 17 and 18 for references.

Project 1 assigned: HMM¨CConstruction, Training and Recognition

Download test data file and labeling file.

Download data file for Viterbi state estimation and the related labeling file.

Due date: Tuesday, March 30, 2004
02/17/2004 Sequential Monte Carlo (II), Assigned papers
02/19/2004 Sequential Monte Carlo (III) Project 2: SMC exercise, Due date: Thursday, April 1, 2004
02/26/2004 Projective Geometry (I)
03/04/2004 Projective Geometry (II)
03/11/2004 Projective Geometry (III)
03/23/2004 Projective Geometry (IV)
03/25/2004 Midterm
03/30/2004 Structure-from-motion, Assigned papers, SfM data
04/01/2004 Edge Detection (I) Assigned papers
04/06/2004 Edge Detection (II) Sample images can be found here.