Masked Generative Motion Planning with Geometry-Guided Token Search
Generative motion planners typically use learned trajectory priors for initial generation, while leaving test-time repair to local continuous refinement. We introduce Masked Generative Motion Planning (MGMP), which extends the learned prior from efficient parallel generation to structural repair. A masked generative tr...
Li-Peng Zhuang, Ying-Dong Ru, Shi-Yu Fan et al.
· 0 citations