Neuromorphic computing, which is inspired by brain neural mechanisms and cognitive behavior mechanisms by means of computational modeling, is emerging with the promise of transforming information processing technology. By learning from the way that the brain processes information in any species, neuromorphic computing includes the simulation and application of the neuronal model of the brain, as well as the more complex in-depth simulation of the information processed by the brain to carry out the thinking process of learning, reasoning and decision-making. Unlike traditional artificial neural networks at the level of software algorithms, neuromorphic computing, which is based on neuroscience theories and biological experimental findings, integrates cognitive science and information science, refers to biological neural network models and architectures and uses either novel neuromorphic devices or existing CMOS devices/circuits to simulate the information processing properties of biological neurons and synapses on software or hardware platforms. Although neuromorphic computing has many benefits, including lower hardware costs, high energy efficiency, self-adaptation, self-learning, self-evolution, and high fault tolerance, they are still in the exploratory stage of development.
This research topic focuses on providing different aspects of brain-inspired neuromorphic computing, from training algorithms for spiking neuron networks, to hardware implementation of neuromorphic computing, and the application of neuromorphic computing. Firstly, it aims to promote the development of specific training methods for spike-based neuron models as well as synaptic models. Secondly, it aims to present the investigation of novel neuromorphic devices, including the fabrication process methods, device structures and the realization of bionic behavior. Thirdly, it aims to advance the neuromorphic circuit design methodology and the applications of neuromorphic computing.
Topics of interest include but are not limited to:
- How neuromorphic devices mimic biological neurons or synapses to realize complex learning and memory behaviors.
- Novel process method for the fabrication of neuromorphic devices.
- Novel structures for neuromorphic devices.
- Improving or proposing new training algorithms based on spiking neural networks.
- Effective application of spiking neural networks in new fields.
- Hardware circuits for spiking neural networks.
- Design and simulation of neuromorphic systems.
Neuromorphic computing, which is inspired by brain neural mechanisms and cognitive behavior mechanisms by means of computational modeling, is emerging with the promise of transforming information processing technology. By learning from the way that the brain processes information in any species, neuromorphic computing includes the simulation and application of the neuronal model of the brain, as well as the more complex in-depth simulation of the information processed by the brain to carry out the thinking process of learning, reasoning and decision-making. Unlike traditional artificial neural networks at the level of software algorithms, neuromorphic computing, which is based on neuroscience theories and biological experimental findings, integrates cognitive science and information science, refers to biological neural network models and architectures and uses either novel neuromorphic devices or existing CMOS devices/circuits to simulate the information processing properties of biological neurons and synapses on software or hardware platforms. Although neuromorphic computing has many benefits, including lower hardware costs, high energy efficiency, self-adaptation, self-learning, self-evolution, and high fault tolerance, they are still in the exploratory stage of development.
This research topic focuses on providing different aspects of brain-inspired neuromorphic computing, from training algorithms for spiking neuron networks, to hardware implementation of neuromorphic computing, and the application of neuromorphic computing. Firstly, it aims to promote the development of specific training methods for spike-based neuron models as well as synaptic models. Secondly, it aims to present the investigation of novel neuromorphic devices, including the fabrication process methods, device structures and the realization of bionic behavior. Thirdly, it aims to advance the neuromorphic circuit design methodology and the applications of neuromorphic computing.
Topics of interest include but are not limited to:
- How neuromorphic devices mimic biological neurons or synapses to realize complex learning and memory behaviors.
- Novel process method for the fabrication of neuromorphic devices.
- Novel structures for neuromorphic devices.
- Improving or proposing new training algorithms based on spiking neural networks.
- Effective application of spiking neural networks in new fields.
- Hardware circuits for spiking neural networks.
- Design and simulation of neuromorphic systems.