eSports Broadcasts with Event Cameras
摘要
Event cameras, characterized by their ability to capture brightness changes through asynchronous events, hold promise for revolutionizing eSports broadcasting. Unlike traditional cameras, they offer high temporal resolution, immense dynamic range, and no motion blur-features crucial for recording the rapid, fine motor skills of eSports athletes. This paper introduces a novel application of event cameras in an eSports environment to enhance the viewer experience by capturing detailed hand movements. We have developed an algorithm to efficiently extract regions of interest from the event data, pinpointing rapid hand movements. This enables directors to effectively highlight the subtle maneuvers of game players, optimizing directorial decisions and enhancing viewer engagement. Additionally, we have collected a dataset of eSports competition scenes for a detailed comparative analysis against standard RGB cameras. Our findings highlight the superior ability of event cameras to capture nuanced motions that are often missed in traditional broadcasts. By utilizing advanced imaging technology, we pave the way for eSports broadcasting that is both high speed and high dynamic range.