TinyNeRF: Towards 100 times Compression of Volume Radiance Fields
文献类型:会议论文
作者 | Zhao TL(赵天理)1,3![]() ![]() ![]() |
出版日期 | 2023-06 |
会议日期 | 2023-02 |
会议地点 | 线上 |
关键词 | Neural Radiance Fields Discrete Cosine Transformation Frequency Domain |
英文摘要 | 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. |
语种 | 英语 |
源URL | [http://ir.ia.ac.cn/handle/173211/52089] ![]() |
专题 | 类脑芯片与系统研究 |
通讯作者 | Cheng J(程健) |
作者单位 | 1.中科院自动化所 2.东南大学 3.中国科学院大学 |
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]. 见:. 线上. 2023-02. |
入库方式: OAI收割
来源:自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。