On November 13, a joint research team led by Professor Yoo Ho-cheon of the Department of Electronic Engineering at Hanyang 토토사이트 db 해킹, along with Professor Kim Yoon-hee's team from Gyeongsang National 토토사이트 db 해킹, Professor Lee Choong-hwan's team from Gachon 토토사이트 db 해킹, and Professor Oh Se-yong's team from Hanyang 토토사이트 db 해킹 ERICA announced that they have developed a 'Spectrally Tuned Floating-Gate Synapse(STFGS)' capable of learning by distinguishing between blue(455nm) and red(660nm) light.

Existing electrical signal-토토사이트 db 해킹 neural devices can only learn through electrical stimulation, leading to high power consumption and difficulty in processing multiple signals 토토사이트 db 해킹 on light wavelength. To overcome this limitation, the research team implemented a dual-wavelength responsive synapse device using two types of organic semiconductors that selectively respond to different wavelengths. Specifically, the device was designed to store photocharges within the floating-gate for an extended period, stably reproducing the Long-Term Potentiation(LTP) phenomenon solely through light stimulation.

Furthermore, the research team designed a wavelength-selective artificial neural network(ANN) 토토사이트 db 해킹 on the STFGS device and achieved an image recognition accuracy of up to 91.37% under 660nm light stimulation conditions. This result demonstrates the potential of a visual neuromorphic device that self-learns by adjusting weights according to optical stimulation conditions.

Professor Yoo Ho-cheon stated, "This research combines the wavelength selectivity of organic semiconductors with a floating-gate structure to implement neural synaptic operation that self-learns based on light stimulation. It can be applied in the future to AI image sensors, wearable visual recognition devices, and low-power AI chips."

This study was supported by the Ministry of Science and ICT and the findings were published online in the international journal Advanced Functional Materials in November 2025. The corresponding paper, Spectrally Tuned Floating-Gate Synapse 토토사이트 db 해킹 on Blue- and Red-Absorbing Organic Molecules for Wavelength-Selective Neural Networks and Fashion Image Classifications, lists Hanyang 토토사이트 db 해킹 Doctoral student Kang Seung-mi as the first author, with Hanyang 토토사이트 db 해킹 Professor Yoo Ho-cheon, Gyeongsang National 토토사이트 db 해킹 Professor Kim Yoon-hee, Gachon 토토사이트 db 해킹 Professor Lee Choong-hwan, and Hanyang 토토사이트 db 해킹 ERICA Professor Oh Se-yong participating as corresponding authors.