Evict3R: Training-Free Token Eviction for Memory-Bounded Streaming Visual Geometry Transformers
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_ | 1866915530993041408 |
|---|---|
| author | Mahdi, Soroush Ayar, Fardin Javanmardi, Ehsan Tsukada, Manabu Javanmardi, Mahdi |
| author_facet | Mahdi, Soroush Ayar, Fardin Javanmardi, Ehsan Tsukada, Manabu Javanmardi, Mahdi |
| contents | Streaming visual transformers like StreamVGGT achieve strong 3D perception but suffer from unbounded growth of key value (KV) memory, which limits scalability. We propose a training-free, inference-time token eviction policy that bounds memory by discarding redundant tokens while keeping the most informative ones. Our method uses significantly less memory with little to no drop in accuracy: on 7-Scenes with long sequences it reduces peak memory from 18.63 GB to 9.39 GB while accuracy and completeness drop by only 0.003. Under strict memory budgets, eviction enables denser frame sampling, which improves reconstruction accuracy compared to the baseline. Experiments across video depth estimation (Sintel, KITTI), 3D reconstruction (7-Scenes, NRGBD), and camera pose estimation (Sintel, TUM-dynamics) show that our approach closely matches StreamVGGT at a fraction of the memory and makes long-horizon streaming inference more practical. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_17650 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Evict3R: Training-Free Token Eviction for Memory-Bounded Streaming Visual Geometry Transformers Mahdi, Soroush Ayar, Fardin Javanmardi, Ehsan Tsukada, Manabu Javanmardi, Mahdi Computer Vision and Pattern Recognition Streaming visual transformers like StreamVGGT achieve strong 3D perception but suffer from unbounded growth of key value (KV) memory, which limits scalability. We propose a training-free, inference-time token eviction policy that bounds memory by discarding redundant tokens while keeping the most informative ones. Our method uses significantly less memory with little to no drop in accuracy: on 7-Scenes with long sequences it reduces peak memory from 18.63 GB to 9.39 GB while accuracy and completeness drop by only 0.003. Under strict memory budgets, eviction enables denser frame sampling, which improves reconstruction accuracy compared to the baseline. Experiments across video depth estimation (Sintel, KITTI), 3D reconstruction (7-Scenes, NRGBD), and camera pose estimation (Sintel, TUM-dynamics) show that our approach closely matches StreamVGGT at a fraction of the memory and makes long-horizon streaming inference more practical. |
| title | Evict3R: Training-Free Token Eviction for Memory-Bounded Streaming Visual Geometry Transformers |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2509.17650 |