中文并非所有任务都同等量化:视觉几何变换器的Fisher引导量化
ENNot All Tasks Quantize Equally: Fisher-Guided Quantization for Visual Geometry Transformer
前馈3D重建模型(如VGGT)可通过单次前向推理同时预测深度、相机位姿和点云,但十亿级参数导致高内存与计算开销。针对此问题,本文提出后训练量化(PTQ)方法,在不显著损失精度前提下压缩模型,使其更适应设备端部署。
arXiv:2605.15828v2 Announce Type: replace Abstract: Feed-forward 3D reconstruction models, represented by Visual Geometry Grounded Transformer (VGGT), jointly predict multiple visual geometry tasks such as depth estimation, camera pose prediction, and point cloud reconstruction in a single forward pass. They have been widely adopted in 3D vision applications, but their billion-scale parameters bring substantial memory and computation overhead, posing challenges for on-device deployment. Post-Training Quantization (PTQ) is an effective technique to reduce this overhead. Existing PTQ methods for