2DNEURALVISION · 2D materials enabled Wide spectrum image sensors and optical NEURAL network based low power consumption computer VISION

From the late 1950s, Computer Vision (CV), a significant field of artificial intelligence (AI), began to be the tool to realize human intelligence and perception. Today, the global computer vision market has a value of over $11 billion and will reach $19 billion in 2027. Current computer vision systems suffer from high power consumption and limited reliability due to the low amount of information captured and sensor malfunction in adverse conditions such as fog, darkness, and bright sunlight. The 2DNEURALVISION project aims to develop the enabling photonic and electronic integrated circuit components for a novel low-power consumption, any weather, any light computer vision system. These components are a 2DM enhanced wide-spectrum image sensor and optical neural network with enabling 2DM components. Non-toxic materials and new waferscale, CMOS-compatible back-end-of-line integration processes for 2DMs will be developed to show compatibility with high-volume markets. The developments in the 2DNEURALVISION project will result in reliable (any weather, any light), low cost (1000x lower), low power consumption (30x lower), and small form factor (1000x lower) computer vision systems. This will have enormous societal impacts by enabling disruptive improvements in the automotive, AR/VR, service robotic, and mobile device sectors.