Task-Oriented Robot-Human Handovers on Legged Manipulators
The researchers introduce AFT‑Handover, a framework that combines large language model‑driven affordance reasoning with texture‑based affordance transfer to enable zero‑shot, task‑oriented robot‑to‑human handovers. In a controlled user study, 71.43% of participants preferred AFT‑Handover over existing state‑of‑the‑art methods, citing reduced regrasping effort and better perceived task understanding. The system was validated on ETH Zürich’s legged manipulator, showcasing seamless handovers on a mobile platform. The work highlights how semantic reasoning can be fused with visual affordance cues for more intuitive human‑robot collaboration.
DiskChunGS: Large-Scale 3D Gaussian SLAM Through Chunk-Based Memory Management
DiskChunGS introduces a scalable 3D Gaussian splatting SLAM pipeline that overcomes traditional GPU memory constraints by treating scene reconstruction as a spatial streaming problem. The system partitions the environment into discrete chunks, keeping only the currently visible regions in GPU...