Under Distinguished Professor William Kaiser, our group has been exploring embedded Machine Learning through the STMicroelectronics SensorTile.
The Group:
- Kavin Balakrishnan
- Arunan Elamaran
- Ryan Tabrizi
- Keshav Vyas
- Michael Davydov
- Atiksh Rao
We have developed an algorithm that we created on SensorTile's SystemWorkbench that can distinguish between stand still, neutral walking, stair ascent, and stair descent as the four primary gaits. The code in this main branch of the repository is the final version of the algorithm that we developed to classify between the four main gaits.
Creating this final result required much testing using various other resources such as MATLAB and Xcode to analyze the different motions in graphical form and test the strength of our decision tree. We display this work in our Development Repository linked here.
Also during the development stage, we used an early version of this final project for data aqusition to then be plotted using MATLAB code. The code can be found in the "data_acquisition" branch, also linked here.
Our research paper is linked here.