All-In-One Neuromorphic Light-Emitting Memristor for Infrared Image Visualization and Recognition

Neuromorphic vision systems have been proven to have great potential in simulating the fundamental functions of the human visual system and can even be extended to the infrared (IR) region, which is invisible to the human eye. Here, an IR neuromorphic light-emitting memristor (INLEM) is presented, which achieves full-process functionality from IR image sensing and information processing to visible light output within an all-in-one structure. The INLEM dynamically adjusts its current in response to different IR stimuli, successfully simulating the transition from short-term potentiation (STP) to long-term potentiation (LTP) in biological synapses. More importantly, this synaptic behavior is visualized in real-time through visible light without the need for an additional display module. Leveraging this feature, the process by which INLEM enhances the contrast of target information by suppressing IR noise is successfully demonstrated. Furthermore, it is proved that the intelligent image processing based on INLEM improves the efficiency of artificial neural networks in image recognition and classification tasks (in both energy and speed). Therefore, this intelligent neuromorphic light-emitting hardware, which integrates IR sensing, memory processing, and display functions, contributes to the development of integrated optoelectronic systems and Internet of Things applications.