LumiView: A Synthetic Object Centric Dataset with Multiple View and Lighting

Zhong Li¹, Liangchen Song², Zhang Chen¹, Xiangyu Du¹, Lele Chen¹, Junsong Yuan², Yi Xu¹,
¹OPPO US Research Center, ²University at Buffalo

Introduction

We used Blender’s physically based path tracer renderer and rendered 3 textured objects: synthetic face, wood train, and face mask. We set up 5 × 5 camera views on the front hemisphere, set 105 directional light sources around the full sphere, and render at a resolution of 800 × 800 pixels. Each camera differs by 10 degrees and each light source differs by 25 degrees on the sphere. Please download the dataset from this link: BlenderData, and put them in the folder data/BlenderData/.

Dataset structure

Each object has its own folder, and each folder contains the following files:

• cam_data_label.npz: camera parameters labeled in label['imgname'],label['intrinsic'] ,label['R_bcam'],w2c_T = label['T_bcam'],l_dirs = label['light_dir'], please refer to the github codelink for more details.

• xx_xx_xx_xx.png: view and light index images.

root/
|-- FaceBase/
|   |-- xx_xx_xx_xx.png # view and light index
|   |-- cam_data_label.npz
|-- facecover/
|-- toytrain/

Citation

@inproceedings{li2023relitneulf,
  title={Relit-NeuLF: Efficient Novel View Synthesis with Neural 4D Light Field},
  author={Li, Zhong, Song, Liangchen, Chen, Zhang, Du, Xiangyu, Chen, Lele, Yuan, Junsong, Xu, Yi},
  booktitle={Proceedings of the 31th ACM International Conference on Multimedia},
  year={2023}
}