LATIDIA · Robótica

Todo para uno: Destilando modelos de fundamentos visuales para una política visuomotora rápida y eficiente en el uso de la memoria.

arXiv:2611.04485v1 Anuncio Tipo: nuevo Resumen: Los modelos de fundamentos visuales (VFM) proporcionan pistas semánticas como profundidad y segmentación útiles para la manipulación robótica, pero su despliegue directo en el control de bucle cerrado es

La noticia

arXiv:2611.04485v1 Announce Type: new Abstract: Visual Foundation Models (VFMs) provide semantic cues such as depth and segmentation useful for robotic manipulation, but directly deploying them in closed-loop control is impractical due to high computational cost, temporal inconsistency, and distribution shift between static internet images and egocentric robot observations. We propose All-For-One (AFO), a two-stage distillation framework that leverages VFMs strictly as privileged information during training, using downstream robotic data. Following the privileged learning paradigm, AFO decouples semantic representation learning from inference-time policy execution. In the All stage, a teacher policy fuses RGB observations with outputs from multiple VFMs to learn a semantically enriched representation. In the For-One stage, this knowledge