Learning Sidewalk Autopilot from Multi-Scale Imitation with Corrective Behavior Expansion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: He, Honglin, Ma, Yukai, Squicciarini, Brad, Wu, Wayne, Zhou, Bolei
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914416499359744
author He, Honglin
Ma, Yukai
Squicciarini, Brad
Wu, Wayne
Zhou, Bolei
author_facet He, Honglin
Ma, Yukai
Squicciarini, Brad
Wu, Wayne
Zhou, Bolei
contents Sidewalk micromobility is a promising solution for last-mile transportation, but current learning-based control methods struggle in complex urban environments. Imitation learning (IL) learns policies from human demonstrations, yet its reliance on fixed offline data often leads to compounding errors, limited robustness, and poor generalization. To address these challenges, we propose a framework that advances IL through corrective behavior expansion and multi-scale imitation learning. On the data side, we augment teleoperation datasets with diverse corrective behaviors and sensor augmentations to enable the policy to learn to recover from its own mistakes. On the model side, we introduce a multi-scale IL architecture that captures both short-horizon interactive behaviors and long-horizon goal-directed intentions via horizon-based trajectory clustering and hierarchical supervision. Real-world experiments show that our approach significantly improves robustness and generalization in diverse sidewalk scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2603_22527
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Learning Sidewalk Autopilot from Multi-Scale Imitation with Corrective Behavior Expansion
He, Honglin
Ma, Yukai
Squicciarini, Brad
Wu, Wayne
Zhou, Bolei
Robotics
Computer Vision and Pattern Recognition
Sidewalk micromobility is a promising solution for last-mile transportation, but current learning-based control methods struggle in complex urban environments. Imitation learning (IL) learns policies from human demonstrations, yet its reliance on fixed offline data often leads to compounding errors, limited robustness, and poor generalization. To address these challenges, we propose a framework that advances IL through corrective behavior expansion and multi-scale imitation learning. On the data side, we augment teleoperation datasets with diverse corrective behaviors and sensor augmentations to enable the policy to learn to recover from its own mistakes. On the model side, we introduce a multi-scale IL architecture that captures both short-horizon interactive behaviors and long-horizon goal-directed intentions via horizon-based trajectory clustering and hierarchical supervision. Real-world experiments show that our approach significantly improves robustness and generalization in diverse sidewalk scenarios.
title Learning Sidewalk Autopilot from Multi-Scale Imitation with Corrective Behavior Expansion
topic Robotics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.22527