TYC数据集

TYC数据集是由达姆施塔特工业大学电气工程与信息技术系合成生物学中心创建的，专注于理解微结构中细胞实例级语义和运动的高分辨率数据集。该数据集包含105个密集标注的高分辨率明场显微镜图像，总计约19,000个实例掩码，以及261个精选视频片段，共1293张高分辨率显微镜图像，用于促进对细胞运动和形态的无监督理解。TYC数据集在实例注释数量、分辨率、复杂性和捕获设备多样性方面超越了现有数据集。此外，数据集还提供了一种标准化的评估策略，以促进未来工作与该数据集的统一比较，适用于生物医学应用，旨在解决细胞分割和跟踪的挑战。

The TYC Dataset was developed by the Center for Synthetic Biology, Department of Electrical Engineering and Information Technology, Technische Universität Darmstadt. This high-resolution dataset focuses on understanding instance-level semantics and motility of cells in microstructures. It comprises 105 densely annotated high-resolution bright-field microscopy images, totaling approximately 19,000 instance masks, as well as 261 curated video clips containing a total of 1293 high-resolution microscopy images, aimed at advancing unsupervised understanding of cell motility and morphology. The TYC Dataset outperforms existing datasets in terms of the number of instance annotations, image resolution, task complexity, and diversity of capture equipment. Additionally, the dataset provides a standardized evaluation protocol to enable unified comparative benchmarking of future works using this resource, tailored for biomedical applications and designed to address the challenges of cell segmentation and tracking.

搜集汇总

背景与挑战

背景概述

TYC数据集是一个高分辨率明场显微镜图像数据集，包含105个密集标注的图像和261个视频片段，总计约19,000个实例掩码和1,293张图像，用于细胞运动和形态的无监督理解。该数据集在实例数量、分辨率和设备多样性方面超越现有数据集，并提供标准化评估策略，适用于细胞分割和跟踪的生物医学研究。