TONAV: Task-Oriented Navigation and Action-Velocity Chunk Learning for Articulated Object Quadrupedal Mobile Manipulation

SummaryAI generated

On September 7, 2026, arXiv cs.CV released the TONAV method, aimed at solving navigation and motion control problems in four-legged movement of articulated objects. This method adopts a task-oriented navigation strategy and combines motion speed block learning techniques to optimize the motion planning and execution capabilities of four-legged robots when dealing with complex articulated objects.

Related eventsRELATED EVENTS

2026-09-09 05:29OpenAI claims huge maths breakthrough on a famed ‘Millennium Problem’
2026-09-09 06:11OpenAI claims to have solved the Navier-Stokes equation problem, but its originality is questioned by the academic community
2026-09-08 22:00Google DeepMind Maps 9 Billion Possible DNA Variants

Key entitiesKEY ENTITIES

TONAVarXiv

Coverage · reports per dayLANGUAGE SPLIT

Entity relations

SignalsSIGNALS

Keyword heat

arXiv1
TONAV1

All reports (1)SOURCES

↗

A arXiv cs.CV en 2026-09-07 12:00

提出 TONAV 框架，通过整合任务导向导航与动作速度块学习，解决四足移动操作机器人中可达性与操作就绪性之间的差距。该框架首先引入位置 - 速度耦合遥操作框架以捕捉运动动力学并收集平滑演示；其次利用视觉语言推理将高层指令分解为可执行子目标并自适应调整基座；最后通过联合建模关节位置及其时间过渡实现稳定持续接触操作。真实世界实验表明，TONAV 在任务导向导航和完整移动操作中的成功率更高，有效缓解了导航 - 操作差距并提升了连续接触交互性能。