The Open Source Forearm Muscle Contraction Detection System
Providing the prosthetics industry with biometric machine learning training data for free.
Open Muscle – The Open-Source Biometric Sensor Designed to Train Neural Networks
Open muscle is a forearm band that detects muscle contractions based on the hall effect and electromagnetic sensing from piston driven skin contact.
Essentially it is a electromechanical band that listens to the sound of muscle contractions to give neural networks the data they need to detect muscle action group contractions that would normally control some movement in the hand.
Open Muscle was conceived to fulfill a need in the prosthetic industry. That need is training data. We at Open Muscle want to provide a platform to gather biometric muscle data in an organized fashion to be used to train future and current neural networks. Most data that is collected is proprietary and not shared with the general public making the data hard to get access to, expensive, and specialized. Our team is working hard to make these data sets easy to collect, inexpensive, and crowd sourced.
Lend a helping hand by submitting your own anonymous training data today!
.
Inexpensive Construction
The most expensive component is currently the battery for Open Muscle. If purchased in bulk the components can cost less than $20 per bracelet. We provide all of the Gerber and STL files for free on our github but encourage others to modify and improve our design and produce them to their hearts content.
High signal to noise ratio
The hall effect sensor and magnet pairs provide a clear high signal to noise ratio. The electromechanical nature of the design filters out certain frequencies mechanically while amplifying others. Again, it is our hope that with an increased training data size that a lot of the problems would be minimized.
Scalable Cell Components
One of the more difficult problems with prosthetic sensors is the sensor placement. Our Open Muscle band is made in cell components that can be arranged in an array of configurations. The primary bracelet configuration is intended to solve the placement problem by increasing the sensor density and increasing the training dataset.
Why Provide This all for Free
It is our belief that there is a place for people to make a profit from their ideas but that is not our role. Our mission is to increase the availability of machine learning training data, increase the availability of prosthetic software and knowledge base, and decrease the cost of prosthetics for everyone.
This is a solvable problem with our current technology we need only enough descriptive data to feed into machine learning models.
Without the large dataset GPT3 would not be what it is today. Without closed captioning and other text and audio combo multimedia voice recognition would not be what it is today. Without Open Muscle or a similar a data collection device the future of prosthetics will only be a pipe dream.
Help solve this solvable problem by volunteering your time, talents, or voice.
Open Muscle Design
