Low-Light Video Enhancement with Synthetic Event Guidance

Low-light video enhancement (LLVE) is an important yet challenging task with many applications such as photographing and autonomous driving. Unlike single image low-light enhancement, most LLVE methods utilize temporal information from adjacent frames to restore the color remove noise of target frame. However, these algorithms, based on framework multi-frame alignment may produce fusion artifacts when encountering extreme low light or fast motion. In this paper, inspired by latency high dynamic range events, we use synthetic events multiple guide restoration videos. Our method contains three stages: 1) event synthesis 2) fusion, 3) enhancement. framework, design two novel modules (event-image transform event-guided dual branch) for second third stages, respectively. Extensive experiments show that our outperforms existing approaches both real datasets. code will be available at https://gitee.com/mindspore/models/tree/master/research/cv/LLVE-SEG.