Compressing Subject-specific Brain-Computer Interface Models into One Model by Superposition in Hyperdimensional Space

• Published in: Proceedings - Design, Automation, and Test in Europe Conference and Exhibition pp. 246 - 251
• Main Authors: Hersche, Michael, Rupp, Philipp, Benini, Luca, Rahimi, Abbas
• Format: Conference Proceeding
• Language: English
• Published: EDAA 01.03.2020
• Subjects: Brain modeling; Computational modeling; Convolution; Electroencephalography; Feature extraction; Kernel; Noise measurement
• ISSN: 1558-1101
• DOI: 10.23919/DATE48585.2020.9116447

Accurate multiclass classification of electroencephalography (EEG) signals is still a challenging task towards the development of reliable motor imagery brain-computer interfaces (MI-BCIs). Deep learning algorithms have been recently used in this area to deliver a compact and accurate model. Reaching high-level of accuracy requires to store subjects-specific trained models that cannot be achieved with an otherwise compact model trained globally across all subjects. In this paper, we propose a new methodology that closes the gap between these two extreme modeling approaches: we reduce the overall storage requirements by superimposing many subject-specific models into one single model such that it can be reliably decomposed, after retraining, to its constituent models while providing a trade-off between compression ratio and accuracy. Our method makes the use of unexploited capacity of trained models by orthogonalizing parameters in a hyperdimensional space, followed by iterative retraining to compensate noisy decomposition. This method can be applied to various layers of deep inference models. Experimental results on the 4-class BCI competition IV-2a dataset show that our method exploits unutilized capacity for compression and surpasses the accuracy of two state-of-the-art networks: (1) it compresses the smallest network, EEGNet [1], by 1.9×, and increases its accuracy by 2.41% (74.73% vs. 72.32%); (2) using a relatively larger Shallow ConvNet [2], our method achieves 2.95 x compression as well as 1.4% higher accuracy (75.05% vs. 73.59%).