TY - GEN
T1 - In-Hardware Learning of Multilayer Spiking Neural Networks on a Neuromorphic Processor
AU - Shrestha, Amar
AU - Fang, Haowen
AU - Rider, Daniel Patrick
AU - Mei, Zaidao
AU - Qiu, Qinru
N1 - Funding Information:
This work is partially supported by the National Science Foundation I/UCRC ASIC (Alternative Sustainable and Intelligent Computing) Center (CNS-1822165).
Publisher Copyright:
© 2021 IEEE.
PY - 2021/12/5
Y1 - 2021/12/5
N2 - Although widely used in machine learning, backpropagation cannot directly be applied to SNN training and is not feasible on a neuromorphic processor that emulates biological neuron and synapses. This work presents a spike-based backpropagation algorithm with biological plausible local update rules and adapts it to fit the constraint in a neuromorphic hardware. The algorithm is implemented on Intel's Loihi chip enabling low power in-hardware supervised online learning of multilayered SNNs for mobile applications. We test this implementation on MNIST, Fashion-MNIST, CIFAR-10 and MSTAR datasets with promising performance and energy-efficiency, and demonstrate a possibility of incremental online learning with the implementation.
AB - Although widely used in machine learning, backpropagation cannot directly be applied to SNN training and is not feasible on a neuromorphic processor that emulates biological neuron and synapses. This work presents a spike-based backpropagation algorithm with biological plausible local update rules and adapts it to fit the constraint in a neuromorphic hardware. The algorithm is implemented on Intel's Loihi chip enabling low power in-hardware supervised online learning of multilayered SNNs for mobile applications. We test this implementation on MNIST, Fashion-MNIST, CIFAR-10 and MSTAR datasets with promising performance and energy-efficiency, and demonstrate a possibility of incremental online learning with the implementation.
KW - Bio-Inspired Approaches
KW - Neuromorphic Computing
KW - Spiking Neural Networks
KW - Supervised Learning
UR - http://www.scopus.com/inward/record.url?scp=85115124549&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85115124549&partnerID=8YFLogxK
U2 - 10.1109/DAC18074.2021.9586323
DO - 10.1109/DAC18074.2021.9586323
M3 - Conference contribution
AN - SCOPUS:85115124549
T3 - Proceedings - Design Automation Conference
SP - 367
EP - 372
BT - 2021 58th ACM/IEEE Design Automation Conference, DAC 2021
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 58th ACM/IEEE Design Automation Conference, DAC 2021
Y2 - 5 December 2021 through 9 December 2021
ER -