Chongfan Technology
News
24
2026
-
07
Wavelength-Encoded Neuromorphic Inference Based on a Microcavity MoS₂ Photodetector Array
Author:
The teams of Ningmu Zou and Zehua Hu at Nanjing University, together with the team of Li Gao at Nanjing University of Posts and Telecommunications, have proposed a bio-inspired optoelectronic inference architecture based on a Fabry–Perot microcavity‑integrated MoS₂ photodetector array, in which sensing, weighting, and accumulation functions are integrated within each pixel. The wavelength selectivity of the cavity design encodes neural network weights into the spectral domain, while the finite carrier lifetime of MoS₂ enables analog temporal accumulation without the need for external memory. The system achieves test accuracies of 99.6%, 94.8%, and 94.0% on the MNIST, CIFAR‑10, and free‑spoken digit datasets, respectively. Post‑training pruning of the optical Hessian matrix further reduces optical complexity while preserving the system’s robustness. This architecture offers a compact pathway toward wavelength‑aware, in‑sensor neuromorphic inference.
The research findings were published in Nature Communications on July 18, 2026, under the title “Wavelength-encoded neuromorphic inference enabled by microcavity MoS₂ photodetector arrays.”
Figure 1: Conceptual diagram and system architecture of bionic optical neural reasoning
Figure 2: Fabrication and characterization of microcavity-integrated MoS₂ photodetectors.
Figure 3: Optical characterization of the microcavity-integrated MoS₂ photodetector.
Figure 4: Experimental setup of the optical classifier
Figure 5: Results of the MoS₂ photodetector-based optical classifier on three 10-class classification datasets (MNIST, CIFAR-10, FSDD)
Figure 6: Optical Hessian pruning for robust, lightweight MoS₂ photodetector optical classification (FSDD)
Source: Optics World
Thank you for visiting the official website of Chongfan Technology. If you have cooperation intentions or suggestions, please contact us through the following methods, and we will reply as soon as possible, thank you!
Address: Room 403, Building 6, Phase III of R&D, No. 36 Xiyong Avenue, High tech Zone, Chongqing, China.
Telephone: +86-13658337211
E-mail: Sales@cfkeji.net
Website: www.cfkeji.net
Mobile Version
