Muscle-Computer Interfaces (muCIs) – Microsoft Research Project

Categories: Microsoft, Tech Talks, Videos

Microsoft research is not resting and is continuously proving to be innovative. Microsoft Research’s new project concentrates on Muscle Based input to computers, which typically uses the electrical signals that our forearms produce. The electrical signals are read by a special device attached to the person’s arms. These signals are compared with the gestures stored and the computer is made to react accordingly!

This project seems to be pretty interesting and I think it will revolutionize the way people interact with their lovely computers. The current model in research provides Six Electromyography Sensors (EMG) placed on a person’s upper right forearms, which serves as the input. Despite these, 2 other additional sensors are also provided at the person’s upper left forearm to recognize person’s hand squeezes.

Event though the detection may be approximately good, the Software must be trained before it can be used. The algorithm used now is the standard Machine learning algorithm, which uses three aspects of EMG:

1. Magnitude of the person’s activity

2. Wave patterns

3. Rate of the activity of muscle

It is also told that once the software knows or understands your gesture, it can determine approx. 80% of your input commands. And this really seems to be pretty cool! The team consisting 7 great people, says that their aim is to provide seamless interaction between the user and the computer with ease. And I really wish that this is going to go big in a very little time. The following video shows a small demo of the research project: