首页|基于NeRF的3D场景动画风格迁移及其技术美学价值研究

基于NeRF的3D场景动画风格迁移及其技术美学价值研究

扫码查看
针对传统3D 动画场景建模技术存在操作困难,相关行业对3D 模型需求量大、计算机硬件要求较高等诸多问题,本文提出一种改进的神经辐射场(NeRF)技术,通过对拍摄影像进行3D重构,进而完成动画风格迁移,借此形成独特3D动画场景艺术效果的解决方案.作为一种新颖的3D 重建方法,NeRF受到计算机图像合成领域广泛关注.本文将NeRF引入3D 场景动画制作中,在验证本方案可行性与科学性的同时,也为3D 动画场景创建提供新思路,充实技术美学价值.
The realisation and technical aesthetic value for 3D scene animation style transfer based on neural radiance fields technology
In view of the difficulties in the operation of traditional 3D animation scene modeling technology,the large de-mand for 3D models in related industries,and high requirements for computer hardwares,this paper proposes an improved neural radiance fields technology,which reconstructs the shot image in three dimensional,and then completes the anima-tion style transfer,thereby forms a unique solution for the technical artistic effect of 3D animation scenes.As a neoteric 3D reconstruction method,neural radiation fields technology has attracted wide attention in the field of computer image synthe-sis.Therefore,this paper introduces neural radiation fields technology into 3D scene animation production,which not only validates the feasibility and scientificity of this scheme,but also provides new ideas for the creation of 3D animation scenes,and thus enriches the aesthetic value of technology.

Neural Radiance FieldsStyle TransferArtificial Intelligence3D Animation Scene

王腾飞、于冰、黄东晋、范正辉

展开 >

上海政法学院上海纪录片学院,上海 201701

上海大学上海电影学院,上海 200072

NeRF 风格迁移 人工智能 3D动画场景

2022年度上海政法学院青年基金

2022XQN18

2024

现代电影技术
广电总局电影技术质量检测所

现代电影技术

CSTPCD
影响因子:0.149
ISSN:1673-3215
年,卷(期):2024.(2)
  • 19