the Sintel 4D Light Field Videos dataset
收藏资源简介:
The generated dataset consists of 23 synthetic 4D LFVs with 1,204x436 pixels, 9x9 views, and 20--50 frames, and has ground-truth depth values in the central view, so that can be used for training deep learning-based methods. Each scene was rendered with a ``clean'' pass after modifying the production file of ``Sintel'' with reference to the MPI Sintel dataset ~\cite{butler2012naturalistic}. A ``clean'' pass includes complex illumination and reflectance properties including specular reflections, such as smooth shading and specular reflections, while bokeh, motion blur, and semi-transparent objects are excluded.
本生成数据集包含23个合成四维光场视频(4D LFVs),分辨率为1204×436像素,包含9×9个视角,帧数量介于20至50帧之间,且中心视角附带真值深度标注,可用于训练基于深度学习的方法。每个场景均参考MPI Sintel数据集~cite{butler2012naturalistic},对《辛特尔》(Sintel)的制作文件进行修改后,通过"纯净渲染通道(clean pass)"完成渲染。该通道涵盖复杂的光照与反射特性,包含镜面反射效果,具体表现为平滑着色与镜面反射,但剔除了散景、运动模糊以及半透明物体。




