MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908353345617920 |
|---|---|
| author | Zhang, Junyi Herrmann, Charles Hur, Junhwa Jampani, Varun Darrell, Trevor Cole, Forrester Sun, Deqing Yang, Ming-Hsuan |
| author_facet | Zhang, Junyi Herrmann, Charles Hur, Junhwa Jampani, Varun Darrell, Trevor Cole, Forrester Sun, Deqing Yang, Ming-Hsuan |
| contents | Estimating geometry from dynamic scenes, where objects move and deform over time, remains a core challenge in computer vision. Current approaches often rely on multi-stage pipelines or global optimizations that decompose the problem into subtasks, like depth and flow, leading to complex systems prone to errors. In this paper, we present Motion DUSt3R (MonST3R), a novel geometry-first approach that directly estimates per-timestep geometry from dynamic scenes. Our key insight is that by simply estimating a pointmap for each timestep, we can effectively adapt DUST3R's representation, previously only used for static scenes, to dynamic scenes. However, this approach presents a significant challenge: the scarcity of suitable training data, namely dynamic, posed videos with depth labels. Despite this, we show that by posing the problem as a fine-tuning task, identifying several suitable datasets, and strategically training the model on this limited data, we can surprisingly enable the model to handle dynamics, even without an explicit motion representation. Based on this, we introduce new optimizations for several downstream video-specific tasks and demonstrate strong performance on video depth and camera pose estimation, outperforming prior work in terms of robustness and efficiency. Moreover, MonST3R shows promising results for primarily feed-forward 4D reconstruction. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_03825 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion Zhang, Junyi Herrmann, Charles Hur, Junhwa Jampani, Varun Darrell, Trevor Cole, Forrester Sun, Deqing Yang, Ming-Hsuan Computer Vision and Pattern Recognition Estimating geometry from dynamic scenes, where objects move and deform over time, remains a core challenge in computer vision. Current approaches often rely on multi-stage pipelines or global optimizations that decompose the problem into subtasks, like depth and flow, leading to complex systems prone to errors. In this paper, we present Motion DUSt3R (MonST3R), a novel geometry-first approach that directly estimates per-timestep geometry from dynamic scenes. Our key insight is that by simply estimating a pointmap for each timestep, we can effectively adapt DUST3R's representation, previously only used for static scenes, to dynamic scenes. However, this approach presents a significant challenge: the scarcity of suitable training data, namely dynamic, posed videos with depth labels. Despite this, we show that by posing the problem as a fine-tuning task, identifying several suitable datasets, and strategically training the model on this limited data, we can surprisingly enable the model to handle dynamics, even without an explicit motion representation. Based on this, we introduce new optimizations for several downstream video-specific tasks and demonstrate strong performance on video depth and camera pose estimation, outperforming prior work in terms of robustness and efficiency. Moreover, MonST3R shows promising results for primarily feed-forward 4D reconstruction. |
| title | MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2410.03825 |