Snap-Snap: Taking Two Images to Reconstruct 3D Human Gaussians in Milliseconds
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909745539973120 |
|---|---|
| author | Lu, Jia Yi, Taoran Fang, Jiemin Yang, Chen Wu, Chuiyun Shen, Wei Liu, Wenyu Tian, Qi Wang, Xinggang |
| author_facet | Lu, Jia Yi, Taoran Fang, Jiemin Yang, Chen Wu, Chuiyun Shen, Wei Liu, Wenyu Tian, Qi Wang, Xinggang |
| contents | Reconstructing 3D human bodies from sparse views has been an appealing topic, which is crucial to broader the related applications. In this paper, we propose a quite challenging but valuable task to reconstruct the human body from only two images, i.e., the front and back view, which can largely lower the barrier for users to create their own 3D digital humans. The main challenges lie in the difficulty of building 3D consistency and recovering missing information from the highly sparse input. We redesign a geometry reconstruction model based on foundation reconstruction models to predict consistent point clouds even input images have scarce overlaps with extensive human data training. Furthermore, an enhancement algorithm is applied to supplement the missing color information, and then the complete human point clouds with colors can be obtained, which are directly transformed into 3D Gaussians for better rendering quality. Experiments show that our method can reconstruct the entire human in 190 ms on a single NVIDIA RTX 4090, with two images at a resolution of 1024x1024, demonstrating state-of-the-art performance on the THuman2.0 and cross-domain datasets. Additionally, our method can complete human reconstruction even with images captured by low-cost mobile devices, reducing the requirements for data collection. Demos and code are available at https://hustvl.github.io/Snap-Snap/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_14892 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Snap-Snap: Taking Two Images to Reconstruct 3D Human Gaussians in Milliseconds Lu, Jia Yi, Taoran Fang, Jiemin Yang, Chen Wu, Chuiyun Shen, Wei Liu, Wenyu Tian, Qi Wang, Xinggang Graphics Computer Vision and Pattern Recognition Reconstructing 3D human bodies from sparse views has been an appealing topic, which is crucial to broader the related applications. In this paper, we propose a quite challenging but valuable task to reconstruct the human body from only two images, i.e., the front and back view, which can largely lower the barrier for users to create their own 3D digital humans. The main challenges lie in the difficulty of building 3D consistency and recovering missing information from the highly sparse input. We redesign a geometry reconstruction model based on foundation reconstruction models to predict consistent point clouds even input images have scarce overlaps with extensive human data training. Furthermore, an enhancement algorithm is applied to supplement the missing color information, and then the complete human point clouds with colors can be obtained, which are directly transformed into 3D Gaussians for better rendering quality. Experiments show that our method can reconstruct the entire human in 190 ms on a single NVIDIA RTX 4090, with two images at a resolution of 1024x1024, demonstrating state-of-the-art performance on the THuman2.0 and cross-domain datasets. Additionally, our method can complete human reconstruction even with images captured by low-cost mobile devices, reducing the requirements for data collection. Demos and code are available at https://hustvl.github.io/Snap-Snap/. |
| title | Snap-Snap: Taking Two Images to Reconstruct 3D Human Gaussians in Milliseconds |
| topic | Graphics Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2508.14892 |