OVGGT: O(1) Constant-Cost Streaming Visual Geometry Transformer
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914516006076416 |
|---|---|
| author | Lu, Si-Yu Chen, Po-Ting Hsu, Hui-Che Jhong, Sin-Ye Cheng, Wen-Huang Chen, Yung-Yao |
| author_facet | Lu, Si-Yu Chen, Po-Ting Hsu, Hui-Che Jhong, Sin-Ye Cheng, Wen-Huang Chen, Yung-Yao |
| contents | Reconstructing 3D geometry from streaming video requires continuous inference under bounded resources. Recent geometric foundation models achieve impressive reconstruction quality through all-to-all attention, yet their quadratic cost confines them to short, offline sequences. Causal-attention variants such as StreamVGGT enable single-pass streaming but accumulate an ever-growing KV cache, exhausting GPU memory within hundreds of frames and precluding the long-horizon deployment that motivates streaming inference in the first place. We present OVGGT, a training-free framework that bounds both memory and compute to a fixed budget regardless of sequence length. Our approach combines Self-Selective Caching, which leverages FFN residual magnitudes to compress the KV cache while remaining fully compatible with FlashAttention, with Dynamic Anchor Protection, which shields coordinate-critical tokens from eviction to suppress geometric drift over extended trajectories. Extensive experiments on indoor, outdoor, and ultra-long sequence benchmarks demonstrate that OVGGT processes arbitrarily long videos within a constant VRAM envelope while achieving state-of-the-art 3D geometric accuracy. Project page: https://vaisr.github.io/OVGGT/ Code: https://github.com/VAISR/OVGGT |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_05959 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | OVGGT: O(1) Constant-Cost Streaming Visual Geometry Transformer Lu, Si-Yu Chen, Po-Ting Hsu, Hui-Che Jhong, Sin-Ye Cheng, Wen-Huang Chen, Yung-Yao Computer Vision and Pattern Recognition Reconstructing 3D geometry from streaming video requires continuous inference under bounded resources. Recent geometric foundation models achieve impressive reconstruction quality through all-to-all attention, yet their quadratic cost confines them to short, offline sequences. Causal-attention variants such as StreamVGGT enable single-pass streaming but accumulate an ever-growing KV cache, exhausting GPU memory within hundreds of frames and precluding the long-horizon deployment that motivates streaming inference in the first place. We present OVGGT, a training-free framework that bounds both memory and compute to a fixed budget regardless of sequence length. Our approach combines Self-Selective Caching, which leverages FFN residual magnitudes to compress the KV cache while remaining fully compatible with FlashAttention, with Dynamic Anchor Protection, which shields coordinate-critical tokens from eviction to suppress geometric drift over extended trajectories. Extensive experiments on indoor, outdoor, and ultra-long sequence benchmarks demonstrate that OVGGT processes arbitrarily long videos within a constant VRAM envelope while achieving state-of-the-art 3D geometric accuracy. Project page: https://vaisr.github.io/OVGGT/ Code: https://github.com/VAISR/OVGGT |
| title | OVGGT: O(1) Constant-Cost Streaming Visual Geometry Transformer |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.05959 |