Bioinspired Retinomorphic Optoelectronic Transistor for Visual Sensing and Processing

Biological visual systems outperform conventional electronics in tasks such as image recognition and restoration due to their low power consumption and rapid processing. To replicate these capabilities, researchers are developing artificial optoelectronic visual systems. Here, we demonstrate a bioinspired retinomorphic optoelectronic transistor (BROT) based on an ITO/ion-gel architecture for neuromorphic vision applications. The BROT device exhibits stable neuromorphic plasticity under electrical and optical stimulation, supporting key synaptic behaviors including long-term potentiation (LTP), long-term depression (LTD), and binary conductance modulation, which enable in-sensor functions such as image recognition and restoration. When integrated with a Generative Adversarial Network (GAN), the BROT device achieves 97.9% image restoration accuracy and 88.3% image recognition accuracy using an artificial neural network (ANN). These results highlight the BROT platform's potential for next-generation optoelectronic synaptic devices and the advancement of neuromorphic vision and brain-inspired computing.