ReS2 Charge Trapping Synaptic Device for Face Recognition Application

Abstract

Synaptic devices are necessary to meet the growing demand for the smarter and more efficient system. In this work, the anisotropic rhenium disulfide (ReS 2) is used as a channel material to construct a synaptic device and successfully emulate the long-term potentiation/depression behavior. To demonstrate that our device can be used in a large-scale neural network system, 165 pictures from Yale Face database are selected for evaluation, of which 120 pictures are used for artificial neural network (ANN) training, and the remaining 45 pictures are used for ANN testing. A three-layer ANN containing more than 10 5 weights is proposed for the face recognition task. Also 120 continuous modulated conductance states are selected to replace weights in our well-trained ANN. The results show that an excellent recognition rate of 100% is achieved with only 120 conductance states, which proves a high potential of our device in the artificial neural network field.

P‐15: Student Poster: Charge Trap‐Based Synaptic Transistor Employing In‐Ga‐Zn‐O as Channel and Trap Layers for Bio‐Inspired Neuromorphic Computing

“…As shown in Figure 6, 32 discrete conductance states were obtained through consecutive potentiation and depression actions. The ratio of maximum conductance to minimum conductance was 4.7, which is a higher ratio than previously reported charge trap-based synaptic devices [24,25]. The higher ratio enables more efficient learning and higher classification accuracy in neuromorphic network operation [26].…”

An artificial synaptic transistor using an α-In2Se3 van der Waals ferroelectric channel for pattern recognition

“…The higher the number of neurons, the higher the accuracy achieved. Using 256 middle neurons, a linear transformation of the device conductance values is performed so that the conductance range is consistent with the weight range using the following relation 46 C j = AI j + B where C j represents the weight value after the linear transformation and A and B are linear transformation coefficients. In the case of 64 weight states, the linear transformation coefficients were A 1 = 2.68 × 10 6 , A 2 = 7.30 × 10 6 and B 1 = −5.473 and B 2 = −1.5083.…”

Tetraphenylethene Carbothioamide‐Based Organic Stimuli‐Responsive Mechanochromic Memristive Devices with Non‐Volatile Memory and Synaptic Learning Functionalities

“…Such kind of EPSC nature is in good agreement with previously reported inorganic devices. [33,34] One of the important forms of short-term synaptic plasticity is the PFF and it is important to understand synaptic information processing. Figure 6d shows the PPF as a function of the interval between the paired pulses.…”