Current VLA models struggle with dynamic manipulation tasks due to a scarcity of dynamic datasets and a reliance on single-frame observations. We introduce DOMINO, a large-scale benchmark for dynamic ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results