The trained model had a Mean Average Error (MAE) value of approximately 25, equating to a 12% error rate. Despite the challenges, the model was able to make real-time predictions, demonstrating its potential for a significant impact on the prosthetic industry.
Predicted Vs Actual from finger 1 or LASK1. The index finger was the most predicted pattern from the open muscle band. Here you see the Actual data from LASK1 in blue and the predicted output of the LASK1 from the model in orange. These are just preliminary data. Everything is published live on the gGitHub The model is 1.7GB so we are not providing the model but all the code is available to train the model with this exact dataset on the gGitHub From attachment predictions_vs_actuals2.csv
TURFPTAx
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.