MU4003 Multimedia Information Storage and Retrieval Syllabus:

MU4003 Multimedia Information Storage and Retrieval Syllabus – Anna University PG Syllabus Regulation 2021

COURSE OBJECTIVES:

 To introduce the basics of multimedia information storage technology, techniques for analysis, representation and retrieval that is commonly used in industry.
 To compare and contrast information retrieval models and internal mechanisms such as Boolean, Probability, and Vector Space Models.
 To outline the structure of queries and media elements.
 To use of machine learning methods on multimedia collections.
 To critically evaluate Multimedia retrieval system effectiveness and improvement techniques.

UNIT I STORAGE AND PRESENTATION OF MULTIMEDIA

Introduction – Media Types – Media Understanding – Description of Audio, Visual Spectral and Video – Storage Networks, Storage Medium – Multidimensional Data Structures: K-D Trees – Point Quadtrees – The MX-Quadtree – Rtrees – Comparison of Different Data Structures.
Suggested Activities:
 Install openCV and learn the functions which are used for Image retrieval.
Suggested Evaluation Methods:
 Quiz on applications of data structure

UNIT II TEXT AND MUSIC RETRIEVAL

Text Information Retrieval: Information Retrieval System – Catalog and Indexing – Automatic Indexing – Term Clustering – User Search Techniques – Information Visualization – Fundamentals – Instantaneous Features – Intensity – Tonal Analysis – Musical Genre, Similarity and Mood.
Suggested Activities:
 Compute the tf-idf weights for the terms car, auto, insurance, best for each document, using the idf values from Figure.
Doc1 Doc2 Doc3
Car 27 4 24
Auto 3 33 0
Insurance 0 33 29
Best 14 0 17
 Consider the query best car insurance on a fictitious collection with N=1,000,000 documents where the document frequencies of auto, best, car and insurance are respectively 5000, 50000, 10000 and 1000. Compute the cosine similarities between the query vector and each document vector in the collection.
Suggested Evaluation Methods:
 Discussion on applying various tf-idf variant and similarity measurements and comparing the results.

UNIT III IMAGE RETRIEVAL

Content-Based Image Retrieval – Techniques – Feature Extraction – Integration – Similarity – Feature in Indexing – Interactive Retrieval – MPEG-7 Standard.
Suggested Activities:
● Assignment on numerical problems on feature extraction techniques.
Suggested Evaluation Methods:
● Tutorial – MPEG-7 standards.
● Tutorial on the problem of choosing the features to be extracted for a large image collection.

UNIT IV VIDEO RETRIEVAL

Content Based Video Retrieval – Video Parsing – Video Abstraction and Summarization – Video Content Representation, Indexing and Retrieval – Video Browsing Schemes – Example of Video Retrieval Systems.
Suggested Activities:
 External learning – Survey on colour-based tracking and optical flow.
 Practical – Learn any open source database software for database operations.
Suggested Evaluation Methods:
 Demonstration and quiz on the practical exercise and the EL component.

UNIT V RETRIEVAL METRICS AND TRENDS

Average Recall and Average Precision – Harmonic Mean – Evaluation of a Search Engine – Relevance Issue – Kappa Measure – Quality Versus Quantity, Possible Factors Which Influence Outcome of a Search – Grandfield Experimental Study – Introduction To Parallel IR – Distributed IR – Trends and Research Issue.
Suggested Activities:
 External learning – Survey on image and video retrieval processing in a search engine such as Google, Yahoo and Bing.
Suggested Evaluation Methods:
 Group discussion and quiz on EL component.
 Assignment on various metric calculations.

TOTAL: 45 PERIODS

COURSE OUTCOMES:

On completion of the course, the students will be able to:
CO1:Learn the basics of multimedia information storage technology, techniques for analysis, representation and retrieval.
CO2:Compare and contrast information retrieval models and internal mechanisms such as Boolean, Probability, and Vector Space Models.
CO3:Implement the process by exploring the open source tool for Image retrieval and video retrieval.
CO4: Recognize the feasibility of applying machine learning for a particular problem.
CO5: Critically evaluate Multimedia retrieval system effectiveness and improvement techniques.

REFERENCES:

1. Christopher D. Manning, Prabhakar Raghavan, Hinrich Schütze, “Introduction to Information Retrieval” , Cambridge University Press, 2008.
2. Philip K. C. Tse, “Multimedia Information Storage and Retrieval: Techniques and Technologies”, IGI Publishing, 2002.
3. Oge Marques, Borko Furht, “Content-Based Image And Video Retrieval”, Springer, 2002.
4. V.S. Subrahmanian, “Principles of Multimedia Database Systems”, Morgan Kaufmann, 1998.
5. Stefan Rüger, “Multimedia Information Retrieval”, Morgan and Claypool Publishers, 2009.