MIPI 2024 Challenge on Demosaic for HybridEVS Camera: Methods and Results

2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)(2024)

引用 0|浏览43
暂无评分
摘要
The increasing demand for computational photography and imaging on mobileplatforms has led to the widespread development and integration of advancedimage sensors with novel algorithms in camera systems. However, the scarcity ofhigh-quality data for research and the rare opportunity for in-depth exchangeof views from industry and academia constrain the development of mobileintelligent photography and imaging (MIPI). Building on the achievements of theprevious MIPI Workshops held at ECCV 2022 and CVPR 2023, we introduce our thirdMIPI challenge including three tracks focusing on novel image sensors andimaging algorithms. In this paper, we summarize and review the Nighttime FlareRemoval track on MIPI 2024. In total, 170 participants were successfullyregistered, and 14 teams submitted results in the final testing phase. Thedeveloped solutions in this challenge achieved state-of-the-art performance onNighttime Flare Removal. More details of this challenge and the link to thedataset can be found at https://mipi-challenge.org/MIPI2024/.
更多
查看译文
关键词
computational photography,demosaicking,hybridEVS
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要