arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.08764cs.CVcs.RO
收藏

快速准确:使用稀疏深度的视频增强

Fast and Accurate: Video Enhancement using Sparse Depth

University of Rochester(罗切斯特大学);Arm Research(Arm研究院);Rochester Institute of Technology(罗切斯特理工学院)

Yu Feng, Patrick Hansen, Paul N. Whatmough, Guoyu Lu, Yuhao Zhu

摘要:

本文提出了一个通用的框架,以建立快速和准确的视频增强任务,如超分辨率,去模糊和去噪算法。对我们的框架来说,重要的是认识到像素流的准确性,而不是密度,是高质量视频增强所需要的。大多数先前的作品采取相反的方法:他们估计密集(每像素)-但通常不太鲁棒-流,主要使用计算成本高的算法。相反,我们提出了一个轻量级的流量估计算法,它融合了稀疏的点云数据和(甚至稀疏和不太可靠)IMU数据在现代自主智能体估计流量信息。建立在流估计之上,我们展示了一个通用框架,该框架将流以即插即用的方式与不同的任务特定层集成在一起。在我们的框架中构建的算法实现了1.78x - 187.41x的加速比,同时比竞争方法提供了0.42 dB - 6.70 dB的质量改进。

英文摘要:

This paper presents a general framework to build fast and accurate algorithms for video enhancement tasks such as super-resolution, deblurring, and denoising. Essential to our framework is the realization that the accuracy, rather than the density, of pixel flows is what is required for high-quality video enhancement. Most of prior works take the opposite approach: they estimate dense (per-pixel)-but generally less robust-flows, mostly using computationally costly algorithms. Instead, we propose a lightweight flow estimation algorithm; it fuses the sparse point cloud data and (even sparser and less reliable) IMU data available in modern autonomous agents to estimate the flow information. Building on top of the flow estimation, we demonstrate a general framework that integrates the flows in a plug-and-play fashion with different task-specific layers. Algorithms built in our framework achieve 1.78x - 187.41x speedup while providing a 0.42 dB - 6.70 dB quality improvement over competing methods.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑