Depth AnyEvent: A Cross-Modal Distillation Paradigm for Event-Based Monocular Depth Estimation
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_ | 1866908545934426112 |
|---|---|
| author | Bartolomei, Luca Mannocci, Enrico Tosi, Fabio Poggi, Matteo Mattoccia, Stefano |
| author_facet | Bartolomei, Luca Mannocci, Enrico Tosi, Fabio Poggi, Matteo Mattoccia, Stefano |
| contents | Event cameras capture sparse, high-temporal-resolution visual information, making them particularly suitable for challenging environments with high-speed motion and strongly varying lighting conditions. However, the lack of large datasets with dense ground-truth depth annotations hinders learning-based monocular depth estimation from event data. To address this limitation, we propose a cross-modal distillation paradigm to generate dense proxy labels leveraging a Vision Foundation Model (VFM). Our strategy requires an event stream spatially aligned with RGB frames, a simple setup even available off-the-shelf, and exploits the robustness of large-scale VFMs. Additionally, we propose to adapt VFMs, either a vanilla one like Depth Anything v2 (DAv2), or deriving from it a novel recurrent architecture to infer depth from monocular event cameras. We evaluate our approach with synthetic and real-world datasets, demonstrating that i) our cross-modal paradigm achieves competitive performance compared to fully supervised methods without requiring expensive depth annotations, and ii) our VFM-based models achieve state-of-the-art performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_15224 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Depth AnyEvent: A Cross-Modal Distillation Paradigm for Event-Based Monocular Depth Estimation Bartolomei, Luca Mannocci, Enrico Tosi, Fabio Poggi, Matteo Mattoccia, Stefano Computer Vision and Pattern Recognition Event cameras capture sparse, high-temporal-resolution visual information, making them particularly suitable for challenging environments with high-speed motion and strongly varying lighting conditions. However, the lack of large datasets with dense ground-truth depth annotations hinders learning-based monocular depth estimation from event data. To address this limitation, we propose a cross-modal distillation paradigm to generate dense proxy labels leveraging a Vision Foundation Model (VFM). Our strategy requires an event stream spatially aligned with RGB frames, a simple setup even available off-the-shelf, and exploits the robustness of large-scale VFMs. Additionally, we propose to adapt VFMs, either a vanilla one like Depth Anything v2 (DAv2), or deriving from it a novel recurrent architecture to infer depth from monocular event cameras. We evaluate our approach with synthetic and real-world datasets, demonstrating that i) our cross-modal paradigm achieves competitive performance compared to fully supervised methods without requiring expensive depth annotations, and ii) our VFM-based models achieve state-of-the-art performance. |
| title | Depth AnyEvent: A Cross-Modal Distillation Paradigm for Event-Based Monocular Depth Estimation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2509.15224 |