Motion Research Progress Report: 2010-04-24

Questions for Mary (to be done over Skype, with both of us looking at the same piece of code at a time):

• Could I see a demo, even if it's just using one moving body, or even just one moving body part? I need to see how systems are started up, etc.  I would like to see how it should like during the May 21 demo.
• Opinion: could I just port this all over to Java?
• How many sensors did you have for generating training data, validation data, and in performance? (Found the answer to two questions: one sensor during training, two during performance)
• Where are the models that you used? Are they in the ml_training folder? How should I use the models? Are they more appropriate for certain movement than others? We have a demo set up for May 21, and I'm wondering which is the best model to use.
• Could you say something more about the edge detection in movement?
• [Confirm my understanding of how the system works overall with her]
• Are there any missing libraries? (Need to double check myself before asking her that question!)

Questions for Thecla

• How many LMA experts can we use to validate the data that we have

Labels: metacreation, motion research