Roger W. Hsiao
Master Candidate
Language Technology Institute
School of Computer Science
Carnegie Mellon University
wrhsiao(AT)cs.cmu.edu
Biography
Roger Hsiao received his B. Eng. and M. Phil. degree in Computer Science from the Hong Kong University of Science and Technology (HKUST) in 2002 and 2004. From August 2005, He becomes a graduate student of Language Technology Institute, School of Computer Science, Carnegie Mellon University. His research interests include speech recognition, machine learning and pattern recognition. In 2002, He represented HKUST in the ACM International Collegiate Programming Contest and became a coach of HKUST programming team in 2003. He received a merit award in the 5th ACM Postgraduate Research Day organized by ACM Hong Kong Chapter in 2004 for his research on guided discriminative training.
My CV
Publications
- Roger Hsiao. Kernel Eigenspace-based MLLR Adaptation. Master Thesis. (ps.gz)
- Brian Mak, James Kwok, Simon Ho and Roger Hsiao. Embedded Kernel Eigenvoice Speaker Adaptation and Its Implication to Reference Speaker Weighting. SAP.
- Roger Hsiao and Brian Mak. A Comparative Study of Two Kernel Eigenspace-based Speaker Adaptation Methods on Large Vocabulary Continuoues Speech Recognition. INTERSPEECH2005.
- Roger Hsiao and Brian Mak. Kernel Eigenspace-based MLLR Adaptation using Multiple Regression Classes. ICASSP2005. (pdf)
- Brian Mak and Roger Hsiao. Improving Eigenspace-based MLLR Adaptation by Kernel PCA. ICSLP2004. (pdf)
- Roger Hsiao and Brian Mak. Discriminative Feature Transformation by Guided Discriminative Training. ICASSP2004. (pdf)
- Brian Mak, Y. C. Tam and Roger Hsiao. Discriminative Training of Auditory Filters of Different Shapes for Robust Speech Recognition. ICASSP2003. (pdf)
More details about the papers
Activities
First of all... be a volunteer! Certainly, we need your help!!
I've joined the Tsuen Wan, Hong Kong chapter of Caritas Internationalis for about three years now. Caritas Internationalis is a global social service organization that helps the needy in over 200 countries. Within the Tsuen Wan chapter, I am a member of the team No Boundary Volunteer (NBV) , whose work primarily focuses on helping the mentally ill. The team consists of 20-30 people, and it is led by a number of experienced social workers. Our regular activities include paying home visits and hospital visits, encouraging social interactions with the patients and their families and supporting assimilation with the community.
Photos of Caritas-HK Volunteering Ceremony 2005 can be accessed here.
Teaching Courses
- Comp201: Java Programming (Fall 2003)
- Introduction to Java programming.
- Fundamentals include language syntax, object oriented programming, inheritance, polymorphism, exception handling, multithreading.
- Standard libraries for input/output, graphics programming, built-in data structures.
- Application programming interface and foundation class library.
- Comp300Y: Introduction to Machine Learning (Summer 2003)
- Introduction to machine learning.
- Covered decision tree learning, artificial neural networks, kernel methods, Bayesian learning, unsupervised learning, and reinforcement learning.
- Comp201: Java Programming (Spring 2003)
- Comp251: Principles of Programming Languages (Fall 2002)
- Comparative studies of programming languages, programming language concepts and constructs.
- Non-imperative programming paradigms: object-oriented, functional, logic, concurrent programming.
- Basic concepts of program translation and interpretation.
- Storage allocation and run-time organization.
Course description comes from here.
Useful Links