Cityscapes Image Pairs城市景观图像对

Cityscapes数据包含从德国驾驶的车辆拍摄的带有标签的视频。该版本是经过处理的子样本，是Pix2Pix论文的一部分。数据集具有来自原始视频的静止图像，并且语义分割标签与原始图像一起显示在图像中。这是用于语义分割任务的最佳数据集之一。

The Cityscapes dataset comprises labeled videos captured from vehicles operating in Germany. This specific version is a processed subsample that was part of the original Pix2Pix paper. The dataset contains static images extracted from the original videos, with their corresponding semantic segmentation labels co-displayed alongside the original images. It is widely recognized as one of the best datasets for semantic segmentation tasks.

搜集汇总

背景与挑战

背景概述

Cityscapes Image Pairs数据集源自德国驾驶车辆拍摄的带标签视频，经Pix2Pix论文处理为静止图像子样本，其中每张图像均以并列方式呈现原始照片和语义分割标签。该数据集专为语义分割任务设计，包含2975个训练文件和500个验证文件，每个文件为256x512像素的合成图像，是此领域的高质量资源。