DIMO: Diverse 3D Motion Generation for Arbitrary Objects
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912699115372544 |
|---|---|
| author | Mou, Linzhan Lei, Jiahui Wang, Chen Liu, Lingjie Daniilidis, Kostas |
| author_facet | Mou, Linzhan Lei, Jiahui Wang, Chen Liu, Lingjie Daniilidis, Kostas |
| contents | We present DIMO, a generative approach capable of generating diverse 3D motions for arbitrary objects from a single image. The core idea of our work is to leverage the rich priors in well-trained video models to extract the common motion patterns and then embed them into a shared low-dimensional latent space. Specifically, we first generate multiple videos of the same object with diverse motions. We then embed each motion into a latent vector and train a shared motion decoder to learn the distribution of motions represented by a structured and compact motion representation, i.e., neural key point trajectories. The canonical 3D Gaussians are then driven by these key points and fused to model the geometry and appearance. During inference time with learned latent space, we can instantly sample diverse 3D motions in a single-forward pass and support several interesting applications including 3D motion interpolation and language-guided motion generation. Our project page is available at https://linzhanm.github.io/dimo. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_07409 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DIMO: Diverse 3D Motion Generation for Arbitrary Objects Mou, Linzhan Lei, Jiahui Wang, Chen Liu, Lingjie Daniilidis, Kostas Computer Vision and Pattern Recognition We present DIMO, a generative approach capable of generating diverse 3D motions for arbitrary objects from a single image. The core idea of our work is to leverage the rich priors in well-trained video models to extract the common motion patterns and then embed them into a shared low-dimensional latent space. Specifically, we first generate multiple videos of the same object with diverse motions. We then embed each motion into a latent vector and train a shared motion decoder to learn the distribution of motions represented by a structured and compact motion representation, i.e., neural key point trajectories. The canonical 3D Gaussians are then driven by these key points and fused to model the geometry and appearance. During inference time with learned latent space, we can instantly sample diverse 3D motions in a single-forward pass and support several interesting applications including 3D motion interpolation and language-guided motion generation. Our project page is available at https://linzhanm.github.io/dimo. |
| title | DIMO: Diverse 3D Motion Generation for Arbitrary Objects |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2511.07409 |