Neural Puppeteer: Keypoint-Based Neural Rendering of Dynamic Shapes (Datset)

This data entry contains the synthetic data from the Neural Puppeteer ACCV 2022 paper.

This data entry contains a public release of the synthetic data from the Neural Puppeteer paper. The data contains synthetic cow, giraffe and pigeon data. We provide depth maps, RGB images, silhouettes and 3D keypoints. We also provide the camera parameters of the 24 camera views (56 in case of pigeons). The data was rendered in Blender by Ole Johannsen.

本数据集条目包含来自《Neural Puppeteer》（ACCV 2022）论文的合成数据。
本数据集公开发布了《Neural Puppeteer》论文的合成数据，涵盖奶牛、长颈鹿与鸽子三类合成样本。我们提供深度图、RGB图像、剪影图以及三维关键点（3D keypoints），同时配套提供24个相机视角的相机参数（鸽子样本的视角数为56）。所有数据均由Ole Johannsen使用Blender渲染生成。