Event
One iPhone Scan Replaces Manual Demonstration Training for Navigation
1 reports · 1 sources
Zhuoyuan Yu and other authors propose Scan2Instruction and PRPT-KS, using a single brief iPhone scan to replace manual demonstrations, constructing training environments for language-guided navigation policies and achieving scene adaptation. The authors state that Scan2Instruction compiles scans into complete training environments and automatically synthesizes executable navigation segments with instructions. In five real environments, the two designs enabled RealNav to achieve an 85% real-world success rate, compared with 49% for the RealNav baseline post-trained on human demonstrations; in unscanned areas of the same building, RealNav still reached 82%, versus 41% for the baseline. All the above success rates are results reported by the authors.
Report timeline
- Scan2Instruction and PRPT-KS: Demonstration-free adaptation for language-guided navigation from a single scan
The study proposes Scan2Instruction and PRPT-KS, which use a single iPhone scan to build a training environment and adapt a language-guided navigation policy without human demonstrations. Across five real environments, RealNav achieved an 85% real-robot success rate, compared with 49% for a baseline post-trained on human demonstrations; in unscanned areas of the same buildings, it still reached 82%, versus 41% for the baseline.
arXiv cs.RO ·