Journal
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS
Volume 29, Issue 2, Pages -Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JSTQE.2023.3245626
Keywords
Optical deep networks; molybdenum disulfide; saturable absorption; nonlinear mapping
Ask authors/readers for more resources
This study investigates the saturable absorption of MoS2 by continuous wave lasers and demonstrates the capability of MoS2 as an activation unit for nonlinear mapping in optical deep networks (ONNs). A simulation-based fully connected neural network is fabricated for mimicking the operation of ONNs and achieving image classification. The recognition accuracy ranged from 89% to 94%, depending on the morphology of MoS2. This article provides a guideline for the selection of nonlinear units and opens up the possibility of implementing all-optical neural networks.
Despite the significant advancements in photonic computation in recent years, the inadequacy of optical nonlinearities limits the scalability of optical deep networks (ONNs). Molybdenum disulfide (MoS2), with excellent nonlinear properties, is emerging as a promising candidate for nonlinear processing. Here, we investigate the saturable absorption of MoS2 by continuous wave lasers and illustrate the capability of MoS2 as an activation unit for nonlinear mapping in ONNs. Moreover, a simulation-based fully connected neural network is fabricated for mimicking the operation of ONNs and demonstrating image classification. The results show that the recognition accurateness ranged from 89% to 94%, depending on the morphology of MoS2. This article provides a guideline for the selection of nonlinear units and opens up the possibility of implementing all-optical neural networks. [GRAPHICS]
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available