Show simple item record

dc.contributor.authorShanker, Shreejith
dc.date.accessioned2022-06-15T12:47:06Z
dc.date.available2022-06-15T12:47:06Z
dc.date.createdAugust 1-3en
dc.date.issued2022
dc.date.submitted2022en
dc.identifier.citationDaniel Flood, Neethu Robinson and Shanker Shreejith, FPGA-based Deep-Learning Accelerators for Energy Efficient Motor Imagery EEG classification, 2022 IEEE International Conference on Omni-layer Intelligent Systems (COINS), Barcelona, Spain, August 1-3, 2022en
dc.identifier.otherY
dc.identifier.urihttp://hdl.handle.net/2262/99556
dc.description.abstractIn recent years, Deep Learning has emerged as a powerful framework for analysing and decoding bio-signals like Electroencephalography (EEG) with applications in brain computer interfaces (BCI) and motor control. Deep convolutional neural networks have shown to be highly effective in decoding BCI signals for applications like two-class motor imagery decoding. Their deployment in real-time applications, however, requires highly parallel and capable computing platforms like GPUs to achieve high-speed inference, consuming a large amount of energy. In this paper, we explore a custom deep learning accelerator on an off-the-shelf hybrid FPGA device to achieve similar inference performance at a fraction of the energy consumption. We evaluate different optimisations at bit-level, data-path and training using a state-of-the-art deep convolutional neural network as our baseline model to arrive at our custom precision quantised deep learning model, which is implemented using the FINN compiler from Xilinx. The accelerator, deployed on a Xilinx Zynq Ultrascale+ FPGA, achieves a significant reduction in power consumption (≈ 17×), sub 2 ms decoding latency and a near-identical decoding accuracy (statistically insignificant reduction of 2.5% average) as the reported baseline subject-specific classification accuracy on an N (= 54) subject motor imagery EEG (MI-EEG) dataset compared to the Deep CNN model on GPU, making our approach more appealing for low-power real-time BCI applications. Furthermore, this design approach is transferable to other deep learning models reported in BCI research, paving the way for novel applications of real-time portable BCI systems.en
dc.language.isoenen
dc.rightsYen
dc.subjectBrain-Computer Interfacesen
dc.subjectDeep Learningen
dc.subjectField Programmable Gate Arraysen
dc.subjectAI Acceleratorsen
dc.titleFPGA-based Deep-Learning Accelerators for Energy Efficient Motor Imagery EEG classificationen
dc.title.alternative2022 IEEE International Conference on Omni-layer Intelligent Systems (COINS)en
dc.typeConference Paperen
dc.type.supercollectionscholarly_publicationsen
dc.type.supercollectionrefereed_publicationsen
dc.identifier.peoplefinderurlhttp://people.tcd.ie/shankers
dc.identifier.rssinternalid244093
dc.rights.ecaccessrightsopenAccess
dc.relation.sourceM.-H. Lee, O.-Y. Kwon, Y.-J. Kim, H.-K. Kim, Y.-E. Lee, J. Williamson, S. Fazli, and S.-W. Lee, ?EEG dataset and OpenBMI toolbox for three BCI paradigms: an investigation into BCI illiteracy,? GigaScience, vol. 8, no. 5, 2019.en
dc.subject.TCDThemeInternational Developmenten
dc.subject.TCDThemeNeuroscienceen
dc.subject.TCDThemeNext Generation Medical Devicesen
dc.subject.TCDThemeSmart & Sustainable Planeten
dc.subject.TCDTagEEG ANALYSISen
dc.subject.TCDTagEEG BRAIN MAPPINGen
dc.subject.TCDTagReconfigurable Computingen
dc.subject.TCDTagdeep learningen
dc.relation.sourceurihttp://gigadb.org/dataset/100542en
dc.identifier.orcid_id0000-0002-9717-1804
dc.status.accessibleNen


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record