Self-Improving 4D Perception via Self-Distillation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915927826628608 |
|---|---|
| author | Huang, Nan Yu, Pengcheng Zeng, Weijia Rehg, James M. Kanazawa, Angjoo Feng, Haiwen Wang, Qianqian |
| author_facet | Huang, Nan Yu, Pengcheng Zeng, Weijia Rehg, James M. Kanazawa, Angjoo Feng, Haiwen Wang, Qianqian |
| contents | Large-scale multi-view reconstruction models have made remarkable progress, but most existing approaches still rely on fully supervised training with ground-truth 3D/4D annotations. Such annotations are expensive and particularly scarce for dynamic scenes, limiting scalability. We propose SelfEvo, a self-improving framework that continually improves pretrained multi-view reconstruction models using unlabeled videos. SelfEvo introduces a self-distillation scheme using spatiotemporal context asymmetry, enabling self-improvement for learning-based 4D perception without external annotations. We systematically study design choices that make self-improvement effective, including loss signals, forms of asymmetry, and other training strategies. Across eight benchmarks spanning diverse datasets and domains, SelfEvo consistently improves pretrained baselines and generalizes across base models (e.g. VGGT and $π^3$), with significant gains on dynamic scenes. Overall, SelfEvo achieves up to 36.5% relative improvement in video depth estimation and 20.1% in camera estimation, without using any labeled data. Project Page: https://self-evo.github.io/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_08532 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Self-Improving 4D Perception via Self-Distillation Huang, Nan Yu, Pengcheng Zeng, Weijia Rehg, James M. Kanazawa, Angjoo Feng, Haiwen Wang, Qianqian Computer Vision and Pattern Recognition Large-scale multi-view reconstruction models have made remarkable progress, but most existing approaches still rely on fully supervised training with ground-truth 3D/4D annotations. Such annotations are expensive and particularly scarce for dynamic scenes, limiting scalability. We propose SelfEvo, a self-improving framework that continually improves pretrained multi-view reconstruction models using unlabeled videos. SelfEvo introduces a self-distillation scheme using spatiotemporal context asymmetry, enabling self-improvement for learning-based 4D perception without external annotations. We systematically study design choices that make self-improvement effective, including loss signals, forms of asymmetry, and other training strategies. Across eight benchmarks spanning diverse datasets and domains, SelfEvo consistently improves pretrained baselines and generalizes across base models (e.g. VGGT and $π^3$), with significant gains on dynamic scenes. Overall, SelfEvo achieves up to 36.5% relative improvement in video depth estimation and 20.1% in camera estimation, without using any labeled data. Project Page: https://self-evo.github.io/. |
| title | Self-Improving 4D Perception via Self-Distillation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.08532 |