1 paper · 1 filter
Shaojie Ren, Jingwen Sun, Zezhi Liu +4
Most simulation benchmarks for robotic manipulation keep target objects static and treat policy inference as instantaneous. These assumptions simplify evaluation but hide how a mov…