FPGA implementation of deep-learning recurrent neural networks with sub-millisecond real-time latency for BCI-decoding of large-scale neural sensors (104 nodes).
Christopher HeelanArto V. NurmikkoWilson TruccoloPublished in: EMBC (2018)
Keyphrases
- recurrent neural networks
- fpga implementation
- deep learning
- real time
- neural model
- recurrent networks
- neural network
- hardware implementation
- unsupervised learning
- feed forward
- unsupervised feature learning
- signal processing
- hidden layer
- artificial neural networks
- field programmable gate array
- echo state networks
- brain computer interface
- machine learning
- mental models
- transfer function
- image processing algorithms
- low cost
- higher order
- information extraction