Retina-Inspired Color-Cognitive Learning via Chromatically Controllable Mixed Quantum Dot Synaptic Transistor Arrays

Area of Science:

• Optoelectronics
• Artificial Intelligence
• Neuroscience

Background:

• Artificial photonic synapses emulate human visual systems by integrating sensing and memory.
• Mimicking retinal functions like multispectral color perception and nonvolatility is crucial for advanced artificial vision.
• Current research often focuses on monochromatic light, leaving color discrimination a significant challenge.

Purpose of the Study:

• To propose an artificial multispectral color recognition system using heterojunction photosynaptic transistors.
• To achieve full-range visible color recognition and controllable synaptic plasticity for artificial visual intelligence.

Main Methods:

• Fabrication of photosynaptic transistors with ratio-controllable mixed quantum dot (M-QD) photoabsorbers and metal-oxide channels.
• Design of M-QD photoabsorbers with specific red (R), green (G), and blue (B)-QD ratios inspired by biological photoreceptors.
• Modulation of synaptic plasticity via gate bias for nonvolatile-to-volatile memory conversion and chromatic control.

Main Results:

• Demonstrated full-range visible color recognition with high photo-to-electric conversion efficiency.
• Achieved adjustable synaptic plasticity enabling nonvolatile-to-volatile memory conversion for chromatic control.
• Developed a 7x7 pixelated photonic synapse array capable of outstanding color image recognition.

Conclusions:

• The proposed M-QD-based photosynaptic transistors offer a viable solution for artificial multispectral color recognition.
• This technology advances the development of artificial visual systems with enhanced color discrimination and image processing capabilities.
• The demonstrated pixelated array showcases the potential for real-world applications in visual intelligence.