Search papers, labs, and topics across Lattice.
9
11
7
13
Action-derived visual attention can boost robot task success rates by over 28% without relying on external labels.
Saliency-guided augmentation can drastically improve the robustness of behavior cloning models to visual domain shifts without sacrificing in-domain performance.
Visual utility in recommendations isn't static; it fluctuates based on user context and item characteristics, revealing critical insights into effective multimodal fusion.
Users achieved a 23.3% faster task completion rate with a co-embodied robotic hand that seamlessly balances autonomy and control.
Achieving "collect one, get one for free" in data collection, MirrorDuo drastically enhances learning efficiency by leveraging mirrored demonstrations.
Unlock olfactory prediction from raw sensor data: SCENT aligns mass spectra with molecular structure, enabling odor prediction without needing explicit chemical formulas.
Keypoint Imitation Learning leaps ahead of RGB baselines in robotic manipulation, but don't expect it to dethrone diffusion models just yet.
Skip the costly real-world data collection: GraspDreamer uses generated human demonstrations from visual generative models to train robots for zero-shot functional grasping.
Forget probabilistic noise models: this new approach learns a latent space where state distances directly encode transition costs, enabling robust multimodal RL.
