Interpreting Brainwaves with SSL

Project Overview and Role

Currently, I am part of a team working on improving the existing BENDR framework for performing self supervised learning on EEG data. A key limitation of the BENDR approach is that it does not handle different sensory layouts accurately. To address this, my team and I proposed including a graph neural network to reduce the multi-channel EEG data. I am responsible for this aspect of the project. I am reducing the time series of each electrode by 1D convolutions, constructing an adjacency matrix where each vertex is an electrode and each edge links spatially adjacent electrodes, inputting the graph structure and node features into a graph neural network, and then passing the resultant embeddings through a latent space alignment block. This enables robustness across varying sensor layouts and incorporates both temporal and spatial information. My role in this project has provided me with a more profound knowledge of neural networks and their applications. However, more importantly, I have strengthened my skills in literature review. I was required to rely on research skills and prior knowledge to understand technical papers in a field I was unfamiliar with. Through this, I’ve learned that many scientific breakthroughs are a result of finding new approaches to existing ideas, requiring the ability to interpret current literature.

View Code (GitHub)