FABLE – Electromyographic Pattern Recognition for Prosthetic Hand
FABLE (Fingers Activated By Low-cost Electronics) is an open-source project for an electromechanical 3D-printed hand. Through the acquisition of signals generated by the muscles myoelectric contraction next to the elbow, it will determine precise finger movements.
A main challenge for the deployment of FABLE in free-living is how to efficiently monitor the muscular activity for temporal pattern analysis and classification. ‘Interesting’ patterns are spaced-out by a big amount of ‘uninteresting’ ones. Spotting of highly temporally sparse patterns is still a challenge for the state-of-the-art machine learning techniques.
The main tasks will be:
- Experimental method design and data collection
- Pattern recognition algorithm design
- Method validation via machine learning
|Project type||Bachelor/Master Thesis|
|Work distribution||25% Data Collection, 75% Data Analysis|
|Useful knowledge||Signal Processing, Machine learning|