NeuroTechSC

8/24 Week Updates

For the week of August 24, the teams have been working on their part of the projects for two weeks now. At this point, many teams are past the initial introduction stage and are now working to complete their goals.

Hardware Team

This week the hardware team had individual members show their progress with learning the intricacies of the Fusion 360 CAD software. The members had been working on their own 3D models of the headset to gain familiarity with the software. As they highlighted their struggles it was suggested that some collaborate with each other’s designs. For example, one model closely followed the AlterEgo Headset while the other was more of a head brace. The final design will come from a collaboration between the members and the experience they have gained. To finalize the meeting a timeline was discussed. They are on track to have a 3D printed headset sometime next week and are mostly finished with the recordings.

Data Team

The data team’s goal for this week is to set up a pipeline from the hardware team, meaning they will be using the BrainFlow application program interface to convert EMG data text into reading. This will all be done in Python by the pipeline group. The processing group in the data team will be working on taking notes on the Butterworth filter, a low-pass signal processing filter.

Machine Learning Team

The machine learning team has been working on building neural network models, specifically convolutional neural networks and recurring networks. Creating these various models will be used as practice and learning until they receive data from the data team. In the upcoming week, they will be sharing their progress and losses with each network as well as the parameters they achieved and the model architecture they used. Recurrent neural networks and convolutional neural networks are distinct networks yet the core of deep learning. Recurrent neural networks have sequential recurrent connections in the models, whereas convolutional neural networks do not.

- 5 toasts