2 ms·Semantic Abstraction: Scene Understanding from 2D Vision-Language Model (2022)1 points by downboots 9mo ago