SMILE: Smooth Motion for Improved Long-Horizon VLA Execution
技术动态
来源：arXiv cs.RO发布时间待核实
arXiv:2608.29432v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models reduce inference cost by executing multiple actions per call, but longer horizons often degrade accuracy because raw chunks contain jitter and outliers. We introduce SMILE, an architecture-preserving interface that predicts B-spline coefficients and decodes them into smooth action sequences.