Toward monolithic 3D integration of in-sensor neuromorphic vision systems using metal-oxide semiconductors

<h2>Summary</h2> Neuromorphic vision systems can realize more compact, and therefore more efficient, computing architectures by integrating sensing, memory, and computing functions on a single platform. Metal-oxide semiconductors play a central role in this transformation due to their electrical characteristics, low-temperature processability, and structural versatility. This review discusses the use of oxide-based optoelectronic synapses, in-memory computing, and logic circuits. Monolithic three-dimensional (M3D) integration is a key enabler of scalable and energy-efficient architectures, and we propose a roadmap for the development of neuromorphic vision systems, including integrated implementations in near- and in-sensor computing. The growing industrial adoption of oxide platforms highlights their technological maturity and practical readiness for next-generation vision systems.