Got the first draft of the ‘refactored’ version done.
- Add a bibliography maybe
No new work this week, focus on getting a lot of writing done.
Got the layman’s version of the article done.
- Finish Proof
- Upload demo data maybe
No new work this week. Want to implement before writing it up.
- Use Blender to get some ground truth / training data for the random forests.
Human Intent Prediction
Prelim. analysis gave indications that this was actually possible. Need to test concrete hypotheses.
- Smooth the data
- Hypothesis: Changes in posture indicate change in desired velocity, otherwise constant
- Controller is relatively simple if this holds
- Possible to learn ‘online’, from human posture etc.
Assigned first project, has some meat.