Panel Discussion

1
Panel Discussion

• Group members
• Tisha Melia
• Alexander Aleman
• Carlo Santamaria
• Alessandro Eccher
• The most important long-range research goal is
• How to handle multimedia data. This area
• encompasses efforts in finding good algorithm
• for feature extraction, intelligent storing,
• and effective searching.

2
Why is this important?
Why is this important?

• Amount of multimedia data keeps increasing.
• Due to specific natures of the data, we need an
  effective way to handle them.

Steps in achieving this goal

• Extraction method.
• finding algorithm to extract important
  features is non-trivial (data still retains
  characteristics after the extraction)?

3
Steps in achieving this goal (cont)?

• Intelligent storage
• Finding a way to effectively store the data
  (avoiding many I/O operations)?
• Specifying the format it is stored in
  (object)?
• Searching
• Comparing features to determine how close
  certain objects are related.
• Good-performing search algorithm.

4
Example of storage format
5
References

• Yossi Rubner, Leonidas Guibas, Carlo Tomasi. The
  Earth Mover's Distance, Multi-Dimensional
  Scaling, and Color-Based Image Retrieval.
• Zhongmin Li, Zhanwu Yu. Object-Based Storage
  Model for Object-Oriented Database.
• A Bamidele, F W M Stentiford, J Morphett. An
  Attention-Based Approach to Content-Based Image
  Retrieval.