Related Experiment Video
Updated: Jan 11, 2026
Assembly and Characterization of Biomolecular Memristors Consisting of Ion Channel-doped Lipid Membranes
Published on: March 9, 2019
A Robust Biomimetic van der Waals Heterostructure Visual Neuromorphic Device for Multiscale In-Sensor Reservoir
Yinxing Zhang1,2, Gongjie Liu3, Yuzhe Zhang4
1Institute of Photoelectronic Thin Film Devices and Technology, Key Laboratory of Optoelectronic Thin Film Devices and Technology of Tianjin, College of Electronic Information and Optical Engineering, Engineering Research Center of Thin Film Photoelectronic Technology of Ministry of Education, Academy for Advanced Interdisciplinary Studies, Nankai University, Tianjin 300350, China.
This study introduces a novel graphdiyne/MoS2 visual neuromorphic device. This all-two-dimensional material heterostructure enhances visual processing and achieves high accuracy in facial and motion recognition tasks.
Area of Science:
- Materials Science
- Neuroscience
- Computer Science
Background:
- Visual neuromorphic devices are essential for processing complex visual data.
- All-two-dimensional material heterostructures offer solutions to lattice mismatch and interfacial defects in conventional devices.
Purpose of the Study:
- To develop a robust visual neuromorphic device using a graphdiyne/MoS2 heterostructure.
- To explore its capabilities in in-sensor reservoir computing for facial and motion recognition.
Main Methods:
- Fabrication of an all-two-dimensional material heterostructure device based on graphdiyne/MoS2.
- Characterization of the device's memory window, on/off ratio, endurance, and stability.
- Development of a multiscale in-sensor reservoir computing system for data analysis.
Main Results:
- Achieved a nearly 10-fold enhancement in memory window due to graphdiyne's unique properties.
- Demonstrated an ultrahigh on/off ratio (5 × 10^7), 70 cycles endurance, and 4 weeks air stability.
- Attained >90% facial recognition accuracy under noise and 95.46% motion trajectory recognition accuracy.
Conclusions:
- The graphdiyne/MoS2 heterostructure shows significant promise for advanced neuromorphic vision systems.
- The developed in-sensor reservoir computing system enables efficient real-time visual data processing.
- This work presents a viable device codesign strategy for next-generation neuromorphic applications.
