VQVLA: Motion-Aware Vector Quantization with Computation Fusion for VLA Acceleration
Overview
Motion-aware vector quantization and computation fusion for efficient vision-language-action models.
通过运动感知向量量化与计算融合,加速视觉语言动作模型(VLA)。
MICRO, 2026. Public paper and code links will be added when available.
Research by Minnan Pei (裴旻楠) and collaborators · Email