Dissertation Progress

1
Dissertation Progress Status Report

• By R. Bassett
• Last Updated Friday, August 21, 2015

2
Exploring 3 areas of Visual Pattern Recognition

• How well do humans do in grading collectible
  items such as coins. What grading variation
  exists amongst the experts. Subjective
• How well does a trained automated system that
  focuses entirely on technical grading do when
  grading collectible items. Technical
• Does a composite interactive experience of human
  / machine grading produce better results than
  either the human or machine by themselves?
  Subjective / Technical

3
Human Grading Experiment
4
Machine Grading Experiment
5
Collaborative Grading Experiment
6
Since December Meeting

• Completed the development and refinement of the
  web based data collection tools.
• Worked with Pace Graduate Assistant on training
  the machine based system that we developed last
  year with new images.
• Have enlisted the help of 5 experts (there may be
  more) to participate in the experiments, all have
  been through the pilot program.
• Finished rough draft version of CHs 1, 2 3

7
Task Completion Goals for 3/15/03

• Have the experts complete the human grading
  experiment.
• Have the automated coin system process and grade
  the 50 test coins and record the results
• Have experts complete the interactive
  collaborative grading experiment
• Start writing CHs 4 5

8
Task Completion Goals for 4/15/03

• Organize the 3 test sets of data, do some number
  crunching perform statistical analysis
• Complete Chapters 4 5
• Complete CH 1, 2 3 and make any required
  changes
• Clean up complete back/front matter