Cinematic Behavior Transfer via NeRF-based Differentiable Filming
Xuekun Jiang, Anyi Rao, Jingbo Wang, Dahua Lin, Bo Dai
Abstract
Figure 1. Given a film shot, a series of visual continuous frames, containing complex camera movement and character motion, we present an approach that estimates the camera trajectory and character motion in world coordinates. The extracted camera and characters' behavior can be applied to new 2D/3D content through our cinematic transfer pipeline. 2D cinematic transfer aims to substitute the characters in the original shot with new 3D characters while preserving the identical character motion and camera movements. 3D cinematic transfer can apply the movements of characters and cameras to new characters and scenes, providing more flexibility to modify various properties, such as lighting, character motion, and camera trajectory.
Ask about this paper
Your agent reads all of it.
Lune indexed this paper to the last equation, along with the top-tier papers that cite it. Ask a question and the answer quotes them.
Cited by top-tier papers4
- Pulp Motion: Framing-aware multimodal camera and human motion generationRobin Courant, Xi WANG, David Loiseaux, Marc Christie et al.ICLR 2026 · 8 citations
- Baking Gaussian Splatting Into Diffusion Denoiser for Fast and Scalable Single-Stage Image-to-3D Generation and ReconstructionYuanhao Cai, He Zhang, Kai Zhang, Yixun Liang et al.ICCV 2025 · 7 citations
- DanceCamAnimator: Keyframe-Based Controllable 3D Dance Camera SynthesisZixuan Wang, Jiayi Li, Xiaoyu Qin, Shikun Sun et al.ACM MM 2024 · 5 citations
- AKiRa: Augmentation Kit on Rays for Optical Video GenerationXi Wang, Robin Courant, Marc Christie, Vicky KalogeitonCVPR 2025
Builds on13
- Instant neural graphics primitives with a multiresolution hash encodingThomas Müller, Alex Evans, Christoph Schied, Alexander KellerSIGGRAPH 2022 · 4,089 citations
- DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D CamerasZachary Teed, Jia DengNeurIPS 2021 · 1,248 citations
- ViTPose: Simple Vision Transformer Baselines for Human Pose EstimationYufei Xu, Jing Zhang, Qiming Zhang, Dacheng TaoNeurIPS 2022 · 1,105 citations
- GNeRF: GAN-based Neural Radiance Field without Posed CameraQuan Meng, Anpei Chen, Haimin Luo, Minye Wu et al.ICCV 2021 · 222 citations
- ProPainter: Improving Propagation and Transformer for Video InpaintingShangchen Zhou, Chongyi Li, Kelvin C. K. Chan, Chen Change LoyICCV 2023 · 205 citations
Related papers
- Virtual Camera Layout Generation using a Reference VideoJung Eun Yoo, Kwanggyoon Seo, Sanghun Park, Jaedong Kim et al.CHI 2021 · 5 citations
- Example-driven virtual cinematography by learning camera behaviorsHongda Jiang, Bin Wang, Xi Wang, Marc Christie et al.SIGGRAPH 2020 · 52 citations
- JAWS: Just A Wild Shot for Cinematic Transfer in Neural Radiance FieldsXi Wang, Robin Courant, Jinglei Shi, Éric Marchand et al.CVPR 2023
- ActCam: Zero-Shot Joint Camera and 3D Motion Control for Video GenerationOmar El Khalifi, Thomas Rossi, Oscar Fossey, Thibault Fouque et al.SIGGRAPH 2026
- Towards Storytelling Animations: Joint Synthesis of Human and Camera MotionsBoyuan Cheng, Yingjie Xi, Rui He, Jinhe Na et al.CVPR 2026
