HiPolicy: Hierarchical Multi-Frequency Action Chunking for Policy Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Jiyao, Han, Zimu, Wang, Junhan, Wu, Xionghao, Lin, Shihong, Li, Jinzhou, Fan, Hongwei, Wu, Ruihai, Li, Dongjiang, Dong, Hao
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914454078226432
author Zhang, Jiyao
Han, Zimu
Wang, Junhan
Wu, Xionghao
Lin, Shihong
Li, Jinzhou
Fan, Hongwei
Wu, Ruihai
Li, Dongjiang
Dong, Hao
author_facet Zhang, Jiyao
Han, Zimu
Wang, Junhan
Wu, Xionghao
Lin, Shihong
Li, Jinzhou
Fan, Hongwei
Wu, Ruihai
Li, Dongjiang
Dong, Hao
contents Robotic imitation learning faces a fundamental trade-off between modeling long-horizon dependencies and enabling fine-grained closed-loop control. Existing fixed-frequency action chunking approaches struggle to achieve both. Building on this insight, we propose HiPolicy, a hierarchical multi-frequency action chunking framework that jointly predicts action sequences at different frequencies to capture both coarse high-level plans and precise reactive motions. We extract and fuse hierarchical features from history observations aligned to each frequency for multi-frequency chunk generation, and introduce an entropy-guided execution mechanism that adaptively balances long-horizon planning with fine-grained control based on action uncertainty. Experiments on diverse simulated benchmarks and real-world manipulation tasks show that HiPolicy can be seamlessly integrated into existing 2D and 3D generative policies, delivering consistent improvements in performance while significantly enhancing execution efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2604_06067
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle HiPolicy: Hierarchical Multi-Frequency Action Chunking for Policy Learning
Zhang, Jiyao
Han, Zimu
Wang, Junhan
Wu, Xionghao
Lin, Shihong
Li, Jinzhou
Fan, Hongwei
Wu, Ruihai
Li, Dongjiang
Dong, Hao
Robotics
Robotic imitation learning faces a fundamental trade-off between modeling long-horizon dependencies and enabling fine-grained closed-loop control. Existing fixed-frequency action chunking approaches struggle to achieve both. Building on this insight, we propose HiPolicy, a hierarchical multi-frequency action chunking framework that jointly predicts action sequences at different frequencies to capture both coarse high-level plans and precise reactive motions. We extract and fuse hierarchical features from history observations aligned to each frequency for multi-frequency chunk generation, and introduce an entropy-guided execution mechanism that adaptively balances long-horizon planning with fine-grained control based on action uncertainty. Experiments on diverse simulated benchmarks and real-world manipulation tasks show that HiPolicy can be seamlessly integrated into existing 2D and 3D generative policies, delivering consistent improvements in performance while significantly enhancing execution efficiency.
title HiPolicy: Hierarchical Multi-Frequency Action Chunking for Policy Learning
topic Robotics
url https://arxiv.org/abs/2604.06067