arXiv AI By Hongsen Cao, Shanxin Yuan, Mona Jaber, Ahmed Sayed

EcoVGGT: Pinned-Key Inference for Efficient and Streamable Feed-Forward 3D Reconstruction from Long Videos

Read the original on arXiv AI →

The Flow has not summarised this story yet — read it at arXiv AI.

arXiv AI
1d ago

Training-Free Acceleration of Multi-View 3D Transformers via Regime-Aware Attention Approximation

The paper introduces RAWATTA, a training‑free technique that accelerates multi‑view 3D transformers by approximating global attention in 8‑bit precision, selectively retaining 16‑bit precision only where necessary. It identifies three functional roles of transformer layers—averaging, matching, and anchoring—and applies different approximations accordingly, including a faster variant that further simplifies averaging and anchoring layers. On a 12 GB GPU with 50 images, RAWATTA speeds up VGGT while losing at most one point in pose accuracy, and both variants can handle 200 images, surpassing the released code’s memory limits.

By Muhammed Izzet Sezer, Sevval Belkis Dikkaya
arXiv Computer Vision
Oct 22

BlazeVGGT: 13x Faster VGGT at 2,000 Frames via Camera Token without Training

BlazeVGGT is a training‑free, plug‑and‑play module that speeds up the VGGT visual geometry model by sparsifying cross‑view attention. It selects a compact set of complementary views using a greedy determinantal point process, preserving dense per‑frame predictions while eliminating redundant computations. On 2,000‑frame sequences, BlazeVGGT achieves roughly 13× faster inference and improves Chamfer Distance from 0.41 to 0.32, demonstrating a strong accuracy–efficiency trade‑off for point‑cloud reconstruction and camera‑pose estimation.

By Shujun Yao, Junhui Lin, Congying Han, Tiande Guo, Ziwen Liu, Bonan Li
arXiv Computation and Language
Sep 4

Select, Compress, Reinvest: A Controlled Study of Visual-Token Allocation in Long-Video MLLMs

The study investigates how long‑video language models decide which frames to keep, compress, and reuse, testing each decision in isolation across six selection rules, three benchmarks, and two answering models. It finds that selecting frames based on queries yields the biggest performance boost, that halving spatial resolution costs little, and that reallocating saved tokens to more compressed frames can further improve accuracy. The work also highlights the importance of a unified evaluation harness to avoid misleading comparisons.

By Prakhar Khatri
arXiv Computer Vision
Oct 28

GPA-VGGT: Geometry-Preserving Acceleration of VGGT

GPA‑VGGT is a training‑free acceleration technique for the VGGT feed‑forward geometry model. It replaces VGGT’s global attention in early blocks with frame‑wise attention that uses a shared cross‑frame key/value set, and in later blocks it merges only keys and values while preserving queries and restoring attention mass. The method achieves speedups comparable to existing efficiency approaches while maintaining over 99% of VGGT’s pose accuracy on multi‑view geometry tasks.

By Muyeol Sung, Jangwon Kim, Jongwoo Lim