Arctique
收藏资源简介:
Arctique是由马克斯-德尔布吕克分子医学中心和亥姆霍兹成像创建的程序生成病理学数据集,旨在模拟H&E染色的结肠组织图像,用于不确定性量化。数据集包含50,000张渲染图像,具有精确的掩码和噪声标签模拟。创建过程基于Blender的3D场景生成框架,允许用户控制图像和标签的不确定性。Arctique主要应用于复杂多对象环境中的不确定性量化方法的基准测试和方法改进,旨在解决图像分割中的不确定性问题。
Arctique is a procedurally generated histopathology dataset created by the Max Delbrück Center for Molecular Medicine and Helmholtz Imaging. It is designed to simulate H&E-stained colon tissue images for uncertainty quantification. The dataset contains 50,000 rendered images, paired with precise segmentation masks and simulated noisy labels. Its development is based on Blender's 3D scene generation framework, which enables users to control the uncertainty of both the generated images and their corresponding labels. Arctique is primarily used for benchmarking and improving uncertainty quantification methods in complex multi-object environments, with the goal of addressing uncertainty-related challenges in image segmentation.

- 1Arctique: An artificial histopathological dataset unifying realism and controllability for uncertainty quantification马克斯-德尔布吕克分子医学中心和亥姆霍兹成像 · 2024年



