Training-free Task-oriented Grasp Generation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Wang, Jiaming, Liu, Diwen, Chen, Jizhuo, Soh, Harold
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911191434002432
author Wang, Jiaming
Liu, Diwen
Chen, Jizhuo
Soh, Harold
author_facet Wang, Jiaming
Liu, Diwen
Chen, Jizhuo
Soh, Harold
contents This paper presents a training-free pipeline for task-oriented grasp generation that combines pre-trained grasp generation models with vision-language models (VLMs). Unlike traditional approaches that focus solely on stable grasps, our method incorporates task-specific requirements by leveraging the semantic reasoning capabilities of VLMs. We evaluate five querying strategies, each utilizing different visual representations of candidate grasps, and demonstrate significant improvements over a baseline method in both grasp success and task compliance rates, with absolute gains of up to 36.9\% in overall success rate. Our results underline the potential of VLMs to enhance task-oriented manipulation, providing insights for future research in robotic grasping and human-robot interaction.
format Preprint
id arxiv_https___arxiv_org_abs_2502_04873
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Training-free Task-oriented Grasp Generation
Wang, Jiaming
Liu, Diwen
Chen, Jizhuo
Soh, Harold
Robotics
This paper presents a training-free pipeline for task-oriented grasp generation that combines pre-trained grasp generation models with vision-language models (VLMs). Unlike traditional approaches that focus solely on stable grasps, our method incorporates task-specific requirements by leveraging the semantic reasoning capabilities of VLMs. We evaluate five querying strategies, each utilizing different visual representations of candidate grasps, and demonstrate significant improvements over a baseline method in both grasp success and task compliance rates, with absolute gains of up to 36.9\% in overall success rate. Our results underline the potential of VLMs to enhance task-oriented manipulation, providing insights for future research in robotic grasping and human-robot interaction.
title Training-free Task-oriented Grasp Generation
topic Robotics
url https://arxiv.org/abs/2502.04873