Styled-QReal: A real-time technique for QR code stylisation

Today QR codes are one of the most widely used scannable patterns, especially in the digital payments ecosystem in India. Traditionally QR codes are made up of randomly appearing black and white square pixels that don’t carry any visual meaning to the normal human being. Hence, this motivates the recent works on beautifying the visual aesthetics of these QR codes. In this paper, we have used the neural style transfer technique to propose an algorithm to generate scannable and robust stylised QR codes in real-time with artistic visual quality, suitable for real-world applications. Also, the experiments performed on the generated QR codes reveal that our QR codes have visual quality at par with existing methods while having better scanning robustness.