Attention Itself Could Retrieve.RetrieveVGGT: Training-Free Long Context Streaming 3D Reconstruction via Query-Key Similarity Retrieval
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_ | 1866909030856785920 |
|---|---|
| author | Zou, Zichen Jia, Xiaosong Wu, Zuxuan Jiang, Yu-Gang |
| author_facet | Zou, Zichen Jia, Xiaosong Wu, Zuxuan Jiang, Yu-Gang |
| contents | Visual Geometry Grounded Transformer (VGGT) advances 3D reconstruction via scalable Transformer architecture, but the quadratic complexity of global attention prevents long context application. StreamVGGT enables streaming with causal attention, yet its KV cache grows linearly with frames, causing memory overflow and quality degradation. We present RetrieveVGGT, a training-free framework, which formulates context construction for VGGT as a retrieval problem. By retrieving a fixed number of relevant frames at each step, VGGT maintains a controllable memory budget, which is close to its training context length. Interestingly, we find that the similarity between current frame queries and cached history frame keys at the first global attention layer of VGGT is already a strong indicator of relevance, eliminating the need for additional learned scoring. To enhance information diversity similar to a recommender system, we propose Segment Sampling so that the retrieval spans distinct relevant segments rather than a single high-similarity region. We design a pose-aware spatial memory mechanism that organizes history frames according to their already estimated camera poses, enabling location-aware retrieval. Extensive experiments demonstrate that RetrieveVGGT achieves state-of-the-art performance, outperforming StreamVGGT, TTT3R, and InfiniteVGGT while maintaining constant memory usage regardless of sequence length. Code is available at https://github.com/zzctmd/RetrieveVGGT. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_09644 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Attention Itself Could Retrieve.RetrieveVGGT: Training-Free Long Context Streaming 3D Reconstruction via Query-Key Similarity Retrieval Zou, Zichen Jia, Xiaosong Wu, Zuxuan Jiang, Yu-Gang Computer Vision and Pattern Recognition Visual Geometry Grounded Transformer (VGGT) advances 3D reconstruction via scalable Transformer architecture, but the quadratic complexity of global attention prevents long context application. StreamVGGT enables streaming with causal attention, yet its KV cache grows linearly with frames, causing memory overflow and quality degradation. We present RetrieveVGGT, a training-free framework, which formulates context construction for VGGT as a retrieval problem. By retrieving a fixed number of relevant frames at each step, VGGT maintains a controllable memory budget, which is close to its training context length. Interestingly, we find that the similarity between current frame queries and cached history frame keys at the first global attention layer of VGGT is already a strong indicator of relevance, eliminating the need for additional learned scoring. To enhance information diversity similar to a recommender system, we propose Segment Sampling so that the retrieval spans distinct relevant segments rather than a single high-similarity region. We design a pose-aware spatial memory mechanism that organizes history frames according to their already estimated camera poses, enabling location-aware retrieval. Extensive experiments demonstrate that RetrieveVGGT achieves state-of-the-art performance, outperforming StreamVGGT, TTT3R, and InfiniteVGGT while maintaining constant memory usage regardless of sequence length. Code is available at https://github.com/zzctmd/RetrieveVGGT. |
| title | Attention Itself Could Retrieve.RetrieveVGGT: Training-Free Long Context Streaming 3D Reconstruction via Query-Key Similarity Retrieval |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.09644 |