arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2301.10241cs.CV
收藏

K-Planes:空间、时间和外观中的显式辐射场

K-Planes: Explicit Radiance Fields in Space, Time, and Appearance

UC Berkeley(加州大学伯克利分校);Istituto Italiano di Tecnologia(意大利技术研究院);Technical University of Denmark(丹麦技术大学)

Sara Fridovich-Keil, Giacomo Meanti, Frederik Warburg, Benjamin Recht, Angjoo Kanazawa

摘要:

我们介绍了k-平面,白盒模型的辐射场在任意尺寸。我们的模型使用d选择2平面来表示d维场景,提供了从静态(d=3)到动态(d=4)场景的无缝方式。这种平面分解使得添加维度特定的先验变得容易,例如时间平滑度和多分辨率空间结构,并且引起场景的静态和动态分量的自然分解。我们使用一个线性特征解码器与学习的颜色基础,产生类似的性能作为非线性黑盒MLP解码器。在一系列合成和真实,静态和动态,固定和变化的外观场景中,k平面可以产生具有竞争力的,通常是最先进的重建保真度,内存使用量低,在完整的4D网格上实现1000倍压缩,并通过纯PyTorch实现快速优化。有关视频结果和代码,请参见https://sarafridov.github.io/K-Planes。

英文摘要:

We introduce k-planes, a white-box model for radiance fields in arbitrary dimensions. Our model uses d choose 2 planes to represent a d-dimensional scene, providing a seamless way to go from static (d=3) to dynamic (d=4) scenes. This planar factorization makes adding dimension-specific priors easy, e.g. temporal smoothness and multi-resolution spatial structure, and induces a natural decomposition of static and dynamic components of a scene. We use a linear feature decoder with a learned color basis that yields similar performance as a nonlinear black-box MLP decoder. Across a range of synthetic and real, static and dynamic, fixed and varying appearance scenes, k-planes yields competitive and often state-of-the-art reconstruction fidelity with low memory usage, achieving 1000x compression over a full 4D grid, and fast optimization with a pure PyTorch implementation. For video results and code, please see https://sarafridov.github.io/K-Planes.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑