Abstract

Time delays are pervasive in neural information processing. To extract motion information in real time, the brain generates anticipative responses to compensate for delays. Here, we show that slow, negative feedback modulation, such as spike frequency adaptation (SFA), or short term depression effect (STP) - a feature widely observed in neural dynamics, enables a neural system to achieve this goal efficiently. We find that SFA or STD enhances the intrinsic mobility of a neural network, manifested by the ability of the network to sustain a spontaneous traveling wave. In response to moving stimuli, the interplay between the external drive and the intrinsic mobility of the network determines the tracking performance of the system, and anticipative tracking occurs when the speed of the traveling wave exceeds that of the external drive. Our study reproduces the anticipative responses of head-direction neurons in spatial navigation, and sheds light on our understanding of how the brain processes motion information in a timely manner.

Furthermore, we show that synapses with short-term depression can be realized by a magnetic tunnel junction, which perfectly reproduces the dynamics of the synaptic weight in a widely applied mathematical model. Then, these dynamical synapses are incorporated into CANNs, which are demonstrated to have the ability to predict a moving object via micromagnetic simulations. This portable spintronics based hardware for neuromorphic computing needs no training and is therefore very promising for the tracking technology for moving targets.

Biography

弭元元，重庆大学医学院神经智能研究中心，研究员。主要专注于大脑在网络层面上处理信息的一般性原理，尤其是神经系统处理动态信息，包括工作记忆、时间节律、运动预测等，的计算机制等。以第一或通讯作者在Neuron, PNAS, Advances in Neural Information Processing Systems (NeurIPS), Front. Comput. Neurosc.，Phys. Rev. E，Europhys. Lett.等杂志发表论文19篇， 在eLife, F1000 Faculty Rev., Nat. Commu.等杂志上合作发表论文20余篇。合作指导的科研课题获得首届全国大学生类脑计算创新应用大赛暨国际邀请赛总决赛一等奖。