Neuromorphic Engineering for Hardware Computational Acceleration and Biomimetic Perception Motion Integration

Abstract

Figure 4. CBRAM-based artificial synapses. a) A conceptual schematic of Ag-SiGe-Si CBRAM during switching. Reproduced with permission. [107] Copyright 2018, Springer Nature. b) Resistive switching characteristics of Ag/CrPS 4 /Au CBRAM. The inset is the schematic diagram of the device. c) The experiment data of LRS retention and the fitting curve of the result with exponential decay function. Reproduced under the terms of the Creative Commons Attribution 4.

Hydrophobic Graphene Quantum Dots for Defect Passivation and Enhanced Moisture Stability of CH3NH3PbI3 Perovskite Solar Cells

“…[ 1–6 ] Rapid progress has been enabled by the outstanding optoelectronic properties of these perovskites, including long charge carrier diffusion lengths, a high optical absorption coefficient, balanced electron and hole mobilities, a long photogenerated carrier lifetime, low trap density, and small exciton binding energies. [ 7–15 ] However, due to low formation energies, the polycrystalline perovskite films contain crystallographic defects and grain boundaries (GBs) at the interface and in the bulk. These defects typically include undercoordinated Pb 2+ ions or halide vacancies on the perovskite surfaces or at the GBs, which create recombination centers and lead to deteriorated performance of the PSCs.…”

Intrinsic polarization coupling in 2D α‐In2Se3toward artificial synapse with multimode operations

“…The synaptic weight increases (long‐term potentiation) when the presynaptic spikes precede the postsynaptic spikes (), whereas synaptic weight decreases (long‐term depression) when the temporal order is reversed (), corresponding to the Hebbian learning rule. On the contrary, the anti‐Hebbian learning rule presents depression for and potentiation for , respectively 46,54,58 . Through designing a specific pulse scheme (Figure S9C), Figure 5B presents a negative synaptic weight change for and vice versa, obeying the asymmetric anti‐Hebbian learning rule.…”

“…48 The human memory is believed to be modulated by the dynamic changes in the strength of the synaptic connections, which is related to the high-order synaptic activities like metaplasticity and learning experience-dependent plasticity. 48,53,54 Thus, the emulation of STP-LTP transition in artificial synaptic devices is essential for the realization of memorizing and forgetting functionalities for the future artificial intelligence systems. Our artificial synaptic device can exhibit two types of conductance states: one undergoes rapid decay after weak signal inputs, analogous to STP, and another one features a long-lived stable state, conceptually analogous to LTP.…”

“…On the contrary, the anti-Hebbian learning rule presents depression for t Δ > 0 and potentiation for t Δ < 0, respectively. 46,54,58 Through designing a specific pulse scheme ( Figure S9C), Figure 5B presents a negative synaptic weight change for t Δ > 0 and vice versa, obeying the asymmetric anti-Hebbian learning rule. The STDP behavior can be quantified by plotting the synaptic weight variation ( W Δ ) as a function of t Δ , and it can be well fitted by the following equations: Figure 5B.…”

“…The LTP can be usually realized by increasing the duration and intensity of the stimulus to reach the threshold value or applying a strong train of stimuli 48 . The human memory is believed to be modulated by the dynamic changes in the strength of the synaptic connections, which is related to the high‐order synaptic activities like metaplasticity and learning experience‐dependent plasticity 48,53,54 . Thus, the emulation of STP–LTP transition in artificial synaptic devices is essential for the realization of memorizing and forgetting functionalities for the future artificial intelligence systems.…”