Quantum Computing Application in Super-Resolution of Degraded Surveillance Imagery for Face Detection with Objective Neural Network
Quantum Computing, Super-Resolution, Face Detection, Neural Network
Super-resolution (SR) is a technique that has been exhaustively exploited as a manner to incorporate strategic possibilities to image processing. As quantum computers gradually advance and provide unconditional proof of a computational advantage at solving certain complex problems over their classical counterparts, the application of quantum computation emerges with the compelling argument to offer exponential speed-up to process computationally expensive operations. Envisioning the design of parallel quantum-ready algorithms and igniting Rapid and Accurate Image Super Resolution (RAISR), an engine for an Objective Neural Network is implemented enhancing degraded surveillance imagery for supporting face detection. This study proposes a series of generative filters to conduct SR, with corresponding assessment regarding image quality, in order to further explorer face detection performance improvements. It presents an extremely efficient method to produce an image that is significantly sharper. Additionally, it illustrates an ultra-sharpening experiment, as a step to induce the learning of more powerful upscaling filters with integrated enhancement effects. As a result, it extends the potential of applying RAISR to improve low quality assets generated by low cost cameras, as well as fosters the eventual implementation of robust face recognition methods powered by SR through the use of quantum computing.