A research team led by Associate Professor Takashi Ikuno from Tokyo University of Science has created a revolutionary artificial synapse that mimics human color vision while generating its own power. Published in Scientific Reports on May 12, 2025, the study demonstrates how this technology could transform machine vision systems in resource-constrained devices.
Unlike conventional optoelectronic systems that require external power sources and substantial computing resources, this self-powered device integrates two different dye-sensitized solar cells that respond uniquely to various wavelengths of light. The synapse exhibits bipolar voltage responses—positive for blue light and negative for red light—enabling it to distinguish colors with remarkable 10-nanometer resolution across the visible spectrum.
This wavelength-dependent behavior allows the device to perform complex logic operations including AND, OR, and XOR within a single component, achieving six-bit resolution with 64 distinct states. When tested in a physical reservoir computing framework, the system successfully classified human movements recorded in different colors with an impressive 82% accuracy rate using just a single synapse, compared to multiple photodiodes required in traditional approaches.
"The results show great potential for the application of this next-generation optoelectronic device to low-power artificial intelligence systems with visual recognition," notes Dr. Ikuno. The technology's applications span multiple industries, from autonomous vehicles that can efficiently recognize traffic signals while conserving battery life to healthcare wearables that monitor vital signs with minimal energy consumption.
By mimicking the human visual system's selective filtering approach rather than processing every detail, this innovation represents a significant step toward bringing sophisticated computer vision capabilities to edge devices like smartphones, drones, and AR/VR systems. The research team envisions this technology contributing to a future where everyday devices can see and interpret the world more like humans do, but with far less energy consumption.