Search papers, labs, and topics across Lattice.
3
0
4
Achieving a 37% improvement in success rates for dexterous manipulation tasks, DexPIE redefines the role of real-world experience in policy training.
Zero-shot robotic manipulation gets a boost: BLaDA grounds open-vocabulary instructions in 3D Gaussian Splatting fields, enabling robots to perform complex dexterous tasks without predefined affordance labels.
Even simple event-based augmentation can rescue vision-language-action models from complete failure in low-light or high-blur scenarios, boosting task success from 0% to 90% in some cases.
