workshop on sparse image and signal processing
TRANSCRIPT
About Sparse Theory and Applications
The sparse representation theory states that most natural signals can be
compactly expressed as a linear combination of a few elementary signals in some
representation matrix. Therefore, sparse representation has become an
invaluable tool as compared to direct time-domain and transform-domain signal
processing methods. Recently, the theory of sparse representation and l1-norm
minimization problem is widely applied in a variety of application areas such as
audio/image/video processing tasks (compression, denoising, deblurring,
inpainting, and superresolution), speech and object recognition, signal detection
and classification, face recognition, array processing, blind source separation,
time-delay estimation, sensor networks, cognitive radios, data acquisition,
solution of partial differential equations, and so on. Each of these applications
demands a design of an efficient and flexible representation matrix, which is
referred to as dictionary in the sparse theory. The sparse representation from
redundant dictionaries may provide better ways to reveal/capture the structures
and also offers better performance in signal modeling and classification problems.
The compressive sensing is a new DSP technique in data acquisition theory.
About CEN The Centre for Excellence in Computational Engineering and Networking (CEN)
is actively involved in research in the board areas of Machine Learning, Image
Processing, Biomedical Signal Processing, Graphics Processing Unit (GPU)
Computing, Software Defined Radio, and Natural Language Processing. The
centre has published more than 120 research papers in reputed peer reviewed
journals and international IEEE conferences. The Center offers two M.Tech
degree programs (Computational Engineering and Networking, and Remote
Sensing and Wireless Sensor Networks). Nine research students are pursuing
PhD degree in the field of Machine Translation for Indian Languages,
Biomedical Signal Processing and Kernel Methods. Many signal processing and
NLP tools have been developed at the Center. Ongoing and completed
research projects at the center are: Machine translation system (MHRD),
Factored SMT (CDAC-Pune), Source Code Plagiarism (DIT), Remote Sensing
Applications (ISRO), Video Summarization and Annotation (ADRIN), SVM-
based Target Classification (NPOL).
First National Workshop on
Sparse Image & Signal Processing (Focus: Sparse Theory, MATLAB & LATEX)
(SISP-2011) 23-26 December, 2011
Organized By
Centre for Excellence in Computational Engineering and Networking
Workshop Highlights
• Introduction to Linear Algebra
• Importance of Mathematics in Engineering
• Signal Transforms (Fourier, DCT, Wavelet)
• Optimization Algorithms
• Sparse Theory and Dictionary Learning
• Compressive Sensing and Imaging
• Single Pixel Camera and 1-bit Sampling
• What are the Open Research Problems?
• Medical Signal and Image Processing
• Hyperspectral Image Analysis
• Speech and Underwater Acoustic Signals
• Sparse Channel Estimation
• Distributed Compressive Video Sensing
• Sparse Image Processing (denoising, deblurring, inpainting, pattern recognition, and compression)
Signal/Image Features and Classifiers
• Feature Extraction Methods
• Supervised and Un-supervised Classifiers
• Importance of Gold Standard Database for Research Works
• Summary of Standard BioSignal , Speech, and Medical and Non-Medical Databases
• Performance Metrics for Validating Algorithms
MATLAB & LATEX Training • Implementation of Algorithms in MATLAB
• GUI Preparation
• Research Ethics
• Training for Preparing a Paper, Thesis , and PPT slides in Latex
Objectives of the Workshop:
The main focus of this workshop is to provide an opportunity for a multidisciplinary group
of academic and industry researchers, scientists, mathematicians, leading engineers,
professionals, scholar students to exchange and share their latest research, and new
ideas in the areas of sparse signal representations, optimization algorithms, and
applications of compressive sensing (CS). The SISP covers theory of sparse recovery and
optimizations, and also gives you a comprehensive understanding of MATALB Toolboxes
(Basic, Signal and Image Processing). The participants will be trained to prepare research
papers, thesis, and PPT slides in Latex.
Registration Form is available @ http://www.nlp.amrita.edu:8080/sisp/
Limited seat is available. Please confirm your participation through mail.
Resource Persons:
Dr. K.R. Ramakrishnan, IISc. Bangalore
Dr. S. R. M. Prasanna, IIT Guwahati
Dr. K. Rajamani, GE Research
Dr. Muralikrishna, NPOL, DRDO, Kochi
Dr. A. Unnikrishnan, NPOL, DRDO, Kochi
Dr. K. P. Soman
Dr. M. Sabarimalai Manikandan
Registration Fee:
• Student: 600
• Faculty: 1000
• Industry: Rs. 1500
Filled up registration form are to be sent
along with the DD for course material fee
drawn in favor of AMRITA CEN payable at
Coimbatore.
Course fee includes:
• Food and Accommodation
• Printed Copy of Text Book on Sparse Signal processing (500 pages)
• Databases (in-house) created for
our research works in the field of
Biometric, Speech, OCR, Transient
Signal Analysis
• Software Packages Created @ CEN
Address for Correspondence:
The coordinator 1
st National Workshop on Sparse Image and Signal Processing
Center for Excellence in Computational Engineering and Networking
Amrita School of Engineering, Amrita Vishwa Vidyapeetham
Amrita Nagar (P.O.), Ettimadai, Coimbatore -641112, Tamilnadu.
Mobile: 9843428800/9994329496/9486221489
E-mail: [email protected]
Last Date for Registration: 15-12-2011