TinyNeRF: Towards 100 times Compression of Volume Radiance Fields

	TinyNeRF: Towards 100 times Compression of Volume Radiance Fields
	Zhao TL(赵天理)1,2 ; Chen JY(陈嘉园)3; Leng C(冷聪)2 ; Cheng J(程健)1,2
	2023-06
会议名称	AAAI Conference on Artificial Intelligence
会议日期	2023-02
会议地点	线上
摘要	Voxel grid representation of 3D scene properties has been widely used to improve the training or rendering speed of the Neural Radiance Fields (NeRF) while at the same time achieving high synthesis quality. However, these methods ac celerate the original NeRF at the expense of extra storage demand, which hinders their applications in many scenar ios. To solve this limitation, we present TinyNeRF, a three stage pipeline: frequency domain transformation, pruning and quantization that work together to reduce the storage demand of the voxel grids with little to no effects on their speed and synthesis quality. Based on the prior knowledge of visual sig nals sparsity in the frequency domain, we convert the origi nal voxel grids in the frequency domain via block-wise dis crete cosine transformation (DCT). Next, we apply pruning and quantization to enforce the DCT coefficients to be sparse and low-bit. Our method can be optimized from scratch in an end-to-end manner, and can typically compress the original models by 2 orders of magnitude with minimal sacrifice on speed and synthesis quality.
关键词	Neural Radiance Fields Discrete Cosine Transformation Frequency Domain
收录类别	EI
语种	英语
是否为代表性论文	是
七大方向——子方向分类	三维视觉
国重实验室规划方向分类	视觉信息处理
是否有论文关联数据集需要存交	否
文献类型	会议论文
条目标识符	http://ir.ia.ac.cn/handle/173211/52089
专题	复杂系统认知与决策实验室_高效智能计算与学习
通讯作者	Cheng J(程健)
作者单位	1.中国科学院大学 2.中科院自动化所 3.东南大学
第一作者单位	中国科学院自动化研究所
通讯作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	Zhao TL,Chen JY,Leng C,et al. TinyNeRF: Towards 100 times Compression of Volume Radiance Fields[C],2023.

条目包含的文件		下载所有文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
TinyNeRF_AAAI.pdf（2855KB）	会议论文		开放获取	CC BY-NC-SA	浏览下载