动态场景下的监督原始视频去噪基准数据集
收藏资源简介:
动态场景下的监督原始视频去噪基准数据集由天津大学电气与信息工程学院创建,包含55组噪声-干净视频对,覆盖ISO值从1600到25600。数据集通过控制对象如玩具的运动,多次捕捉静态瞬间以生成干净视频帧。该数据集是首个包含噪声-干净对的动态视频数据集,旨在解决动态场景下视频去噪的难题,支持利用时间相关性进行去噪,并实现对真实噪声视频的定量评估。
The supervised raw video denoising benchmark dataset for dynamic scenes was created by the School of Electrical and Information Engineering, Tianjin University. It includes 55 pairs of noisy-clean videos, with ISO values ranging from 1600 to 25600. The dataset generates clean video frames by capturing static instants multiple times while controlling the movement of objects such as toys. As the first dynamic video dataset with noisy-clean pairs, it aims to address the challenges of video denoising in dynamic scenes, support denoising by leveraging temporal correlations, and enable quantitative evaluation of real noisy videos.

- 1Supervised Raw Video Denoising with a Benchmark Dataset on Dynamic Scenes天津大学电气与信息工程学院 · 2020年



