RDD: Retrieval-Based Demonstration Decomposer for Planner Alignment in Long-Horizon Tasks
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914096681582592 |
|---|---|
| author | Yan, Mingxuan Wang, Yuping Liu, Zechun Li, Jiachen |
| author_facet | Yan, Mingxuan Wang, Yuping Liu, Zechun Li, Jiachen |
| contents | To tackle long-horizon tasks, recent hierarchical vision-language-action (VLAs) frameworks employ vision-language model (VLM)-based planners to decompose complex manipulation tasks into simpler sub-tasks that low-level visuomotor policies can easily handle. Typically, the VLM planner is finetuned to learn to decompose a target task. This finetuning requires target task demonstrations segmented into sub-tasks by either human annotation or heuristic rules. However, the heuristic subtasks can deviate significantly from the training data of the visuomotor policy, which degrades task performance. To address these issues, we propose a Retrieval-based Demonstration Decomposer (RDD) that automatically decomposes demonstrations into sub-tasks by aligning the visual features of the decomposed sub-task intervals with those from the training data of the low-level visuomotor policies. Our method outperforms the state-of-the-art sub-task decomposer on both simulation and real-world tasks, demonstrating robustness across diverse settings. Code and more results are available at rdd-neurips.github.io. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_14968 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RDD: Retrieval-Based Demonstration Decomposer for Planner Alignment in Long-Horizon Tasks Yan, Mingxuan Wang, Yuping Liu, Zechun Li, Jiachen Robotics Artificial Intelligence Computer Vision and Pattern Recognition Machine Learning Systems and Control To tackle long-horizon tasks, recent hierarchical vision-language-action (VLAs) frameworks employ vision-language model (VLM)-based planners to decompose complex manipulation tasks into simpler sub-tasks that low-level visuomotor policies can easily handle. Typically, the VLM planner is finetuned to learn to decompose a target task. This finetuning requires target task demonstrations segmented into sub-tasks by either human annotation or heuristic rules. However, the heuristic subtasks can deviate significantly from the training data of the visuomotor policy, which degrades task performance. To address these issues, we propose a Retrieval-based Demonstration Decomposer (RDD) that automatically decomposes demonstrations into sub-tasks by aligning the visual features of the decomposed sub-task intervals with those from the training data of the low-level visuomotor policies. Our method outperforms the state-of-the-art sub-task decomposer on both simulation and real-world tasks, demonstrating robustness across diverse settings. Code and more results are available at rdd-neurips.github.io. |
| title | RDD: Retrieval-Based Demonstration Decomposer for Planner Alignment in Long-Horizon Tasks |
| topic | Robotics Artificial Intelligence Computer Vision and Pattern Recognition Machine Learning Systems and Control |
| url | https://arxiv.org/abs/2510.14968 |