@misc{indiciae31caa57648c1,
  title = {FlashVGGT: Efficient and Scalable Visual Geometry Transformers with Compressed Descriptor Attention},
  author = {Zipeng Wang and Dan Xu},
  year = {2026},
  url = {https://arxiv.org/abs/2512.01540},
  note = {Source identifier: 2512.01540}
}