MIT-Adobe FiveK

该数据集名为MIT-Adobe FiveK，包含了由五位专家使用Adobe Lightroom处理的照片，旨在评估图像增强方法。与LOL数据集相比，该数据集规模更大，涵盖了更多样化的场景。总共有5000张图片，其中4500张用于训练，500张用于测试，其任务是进行图像增强效果的评估。

This dataset, named MIT-Adobe FiveK, comprises photographs processed by five experts via Adobe Lightroom, and is designed for evaluating image enhancement methods. Compared with the LOL dataset, it features a larger scale and covers a wider range of diverse scenarios. The dataset contains a total of 5000 images, with 4500 allocated for training and 500 for testing, and its designated task is to assess the performance of image enhancement approaches.

搜集汇总

背景与挑战

背景概述

MIT-Adobe FiveK数据集包含5000张由五位专家处理的图片，主要用于图像增强方法评估，其中4500张用于训练，500张用于测试，覆盖多样化场景。