3DGStream: On-the-Fly Training of 3D Gaussians for Efficient Streaming of Photo-Realistic Free-Viewpoint Videos
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866917690371735552 |
|---|---|
| author | Sun, Jiakai Jiao, Han Li, Guangyuan Zhang, Zhanjie Zhao, Lei Xing, Wei |
| author_facet | Sun, Jiakai Jiao, Han Li, Guangyuan Zhang, Zhanjie Zhao, Lei Xing, Wei |
| contents | Constructing photo-realistic Free-Viewpoint Videos (FVVs) of dynamic scenes from multi-view videos remains a challenging endeavor. Despite the remarkable advancements achieved by current neural rendering techniques, these methods generally require complete video sequences for offline training and are not capable of real-time rendering. To address these constraints, we introduce 3DGStream, a method designed for efficient FVV streaming of real-world dynamic scenes. Our method achieves fast on-the-fly per-frame reconstruction within 12 seconds and real-time rendering at 200 FPS. Specifically, we utilize 3D Gaussians (3DGs) to represent the scene. Instead of the naïve approach of directly optimizing 3DGs per-frame, we employ a compact Neural Transformation Cache (NTC) to model the translations and rotations of 3DGs, markedly reducing the training time and storage required for each FVV frame. Furthermore, we propose an adaptive 3DG addition strategy to handle emerging objects in dynamic scenes. Experiments demonstrate that 3DGStream achieves competitive performance in terms of rendering speed, image quality, training time, and model storage when compared with state-of-the-art methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_01444 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | 3DGStream: On-the-Fly Training of 3D Gaussians for Efficient Streaming of Photo-Realistic Free-Viewpoint Videos Sun, Jiakai Jiao, Han Li, Guangyuan Zhang, Zhanjie Zhao, Lei Xing, Wei Computer Vision and Pattern Recognition Constructing photo-realistic Free-Viewpoint Videos (FVVs) of dynamic scenes from multi-view videos remains a challenging endeavor. Despite the remarkable advancements achieved by current neural rendering techniques, these methods generally require complete video sequences for offline training and are not capable of real-time rendering. To address these constraints, we introduce 3DGStream, a method designed for efficient FVV streaming of real-world dynamic scenes. Our method achieves fast on-the-fly per-frame reconstruction within 12 seconds and real-time rendering at 200 FPS. Specifically, we utilize 3D Gaussians (3DGs) to represent the scene. Instead of the naïve approach of directly optimizing 3DGs per-frame, we employ a compact Neural Transformation Cache (NTC) to model the translations and rotations of 3DGs, markedly reducing the training time and storage required for each FVV frame. Furthermore, we propose an adaptive 3DG addition strategy to handle emerging objects in dynamic scenes. Experiments demonstrate that 3DGStream achieves competitive performance in terms of rendering speed, image quality, training time, and model storage when compared with state-of-the-art methods. |
| title | 3DGStream: On-the-Fly Training of 3D Gaussians for Efficient Streaming of Photo-Realistic Free-Viewpoint Videos |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2403.01444 |