Pasar al contenido principal

BENCHMARKS FOR UNDERSTANDING GRASPING

PCI2019-103447

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos Programación Conjunta Internacional
Año convocatoria 2019
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario AGENCIA ESTATAL CONSEJO SUPERIOR DE INVESTIGACIONES CIENTIFICAS (CSIC)

Resultados relacionados

Resultados totales (Incluyendo duplicados): 6
Encontrada(s) 1 página(s)

DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Contribución de congreso (ConferenceOutput)

Benchmarking cloth manipulation using action graphs: an example in placing flat

  • García-Camacho, Irene
  • Borràs, Julia
  • Alenyà, Guillem
Trabajo presentado en el IROS Workshop on Benchmarking of robotic grasping and manipulation: protocols, metrics and data analysis, celebrado online desde Praga (República Checa), el 27 de septiembre de 2021, Benchmarking robotic manipulation is complex due to the difficulty in reproducing and comparing results across different embodiments and scenarios. Cloth manipula- tion presents additional challenges due to the complex object configuration space. Traditional cloth manipulation papers do not have well defined metrics to evaluate the success of a task or the quality of the result, and are tailored to each evaluation. In this paper we propose to evaluate cloth manipulation seg- menting a task into steps that can be evaluated independently, and to study how their success measures influence in the next segment and relate to task as a whole. In particular, we study a popular task such as placing a cloth flat on a table. We propose a benchmark with simple but continuous evaluation metrics that explore the influence of grasp location into the quality of the task. Our results show that grasp location doesn¿t need to be precise on corners, that quality measures focused on evaluating different cloth parts can enlighten issues to solve and that success definition of a segment has to consider its influence on the ability to perform successfully the next segment of action., This work receives funding from the Spanish State Research Agency through the BURG project (CHIST-ERA - PCIN2019-103447) and the Mar´ıa de Maeztu Seal of Excellence to IRI (MDM-2016-0656).
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Household Cloth Object Set: Fostering Benchmarking in Deformable Object Manipulation

  • García-Camacho, Irene
  • Borràs, Julia
  • Calli, Berk
  • Norton, Adam
  • Alenyà, Guillem
Benchmarking of robotic manipulations is one of the open issues in robotic research. An important factor that has enabled progress in this area in the last decade is the existence of common object sets that have been shared among different research groups. However, the existing object sets are very limited when it comes to cloth-like objects that have unique particularities and challenges. This letter is a first step towards the design of a cloth object set to be distributed among research groups from the robotics cloth manipulation community. We present a set of household cloth objects and related tasks that serve to expose the challenges related to gathering such an object set and propose a roadmap to the design of common benchmarks in cloth manipulation tasks, with the intention to set the grounds for a future debate in the community that will be necessary to foster benchmarking for the manipulation of cloth-like objects. Some RGB-D and object scans are collected as examples for the relevant configurations and shared in http://www.iri.upc.edu/groups/perception/ClothObjectSet/., The work of IGC is partially supported by MCIN/ AEI /10.13039/501100011033 and by the ”European Union” under project BURG (PCI2019-103447); JB thanks MCIN/ AEI /10.13039/501100011033 under project CHLOE-GRAPH (PID2020-118649RB-I00); GA acknowledges the EU H2020 Programme under grant agreement ERC–2016–ADG–741930 (CLOTHILDE); and BC thanks the National Science Foundation under grant IIS-1900953
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Contribución de congreso (ConferenceOutput)

Cloth manipulation and perception competition

  • García-Camacho, Irene
  • Borràs, Julia
  • Calli, Berk
  • Norton, Adam
  • Alenyà, Guillem
Trabajo presentado en el 2nd Workshop on Representing and Manipulating Deformable Objects (ICRA 2022), celebrado online y presencial, del 23 al 31 de mayo de 2022. Disponible en Youtube: https://www.youtube.com/watch?v=EVAXd6waVqg, In the last decade, several competitions in robotic manipulation have been organised as a way to drive scientific progress in the field. They enable comparison of different approaches through a well-defined benchmark with equal test conditions. However, current competitions usually focus on rigid-object manipulation, leaving behind the challenges that suppose grasping deformable objects, especially highly-deformable ones as cloth-like objects. In this paper, we want to present the first competition in perception and manipulation of textile objects as an eficient method to accelerate scientific progress in the domain of domestic service robots. To do so, we selected a small set of tasks to benchmark in a common framework using the same set of objects and assessment methods. This competition has been conceived to freely distribute the Household Cloth Object Set to research groups working on cloth manipulation and perception and participate on the challenge. In this work, we present an overview of the tasks that are proposed in the competition, detailed descriptions of the tasks and more information on the scoring and rules are provided in the website http://www.iri.upc.edu/groups/perception/ClothManipulationChallenge/, This work receives funding by MCIN/ AEI /10.13039/501100011033 and by the ”European Union” under project BURG (PCI2019-103447)
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Garment manipulation dataset for robot learning by demonstration through a virtual reality framework

  • Boix Granell, Arnau
  • Foix, Sergi
  • Torras, Carme
Being able to teach complex capabilities, such as folding garments, to a bi-manual robot is a very challenging task, which is often tackled using learning from demonstration datasets. The few garment folding datasets available nowadays to the robotics research community are either gathered from human demonstrations or generated through simulation. The former have the huge problem of perceiving human action and transferring it to the dynamic control of the robot, while the latter requires coding human motion into the simulator in open loop, resulting in far-from-realistic movements. In this article, we present a reduced but very accurate dataset of human cloth folding demonstrations. The dataset is collected through a novel virtual reality (VR) framework we propose, based on Unity¿s 3D platform and the use of a HTC Vive Pro system. The framework is capable of simulating very realistic garments while allowing users to interact with them, in real time, through handheld controllers. By doing so, and thanks to the immersive experience, our framework gets rid of the gap between the human and robot perception-action loop, while simplifying data capture and resulting in more realistic samples., This work was developed in the context of the project CLOTHILDE (”CLOTH manIpulation Learning from DEmonstrations”) which has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement No. 741930) and is also supported by the BURG project PCI2019-103447 funded by MCIN/ AEI /10.13039/501100011033 and by the ”European Union”.
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Garment manipulation dataset for robot learning by demonstration through a virtual reality framework

  • Boix Granell, Arnau
  • Foix Salmerón, Sergi|||0000-0001-9249-6696
  • Torras, Carme|||0000-0002-2933-398X
., Being able to teach complex capabilities, such as folding garments, to a bi-manual robot is a very challenging task, which is often tackled using learning from demonstration datasets. The few garment folding datasets available nowadays to the robotics research community are either gathered from human demonstrations or generated through simulation. The former have the huge problem of perceiving human action and transferring it to the dynamic control of the robot, while the latter requires coding human motion into the simulator in open loop, resulting in far-from-realistic movements. In this article, we present a reduced but very accurate dataset of human cloth folding demonstrations. The dataset is collected through a novel virtual reality (VR) framework we propose, based on Unity’s 3D platform and the use of a HTC Vive Pro system. The framework is capable of simulating very realistic garments while allowing users to interact with them, in real time, through handheld controllers. By doing so, and thanks to the immersive experience, our framework gets rid of the gap between the human and robot perception-action loop, while simplifying data capture and resulting in more realistic sample, This work was developed in the context of the project CLOTHILDE (”CLOTH manIpulation Learning from DEmonstrations”) which has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement No. 741930) and is also supported by the BURG project PCI2019-103447 funded by MCIN/ AEI /10.13039/501100011033 and by the ”European Union”., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Garment manipulation dataset for robot learning by demonstration through a virtual reality framework

  • Boix Granell, Arnau
  • Foix Salmerón, Sergi
  • Torras, Carme
., Being able to teach complex capabilities, such as folding garments, to a bi-manual robot is a very challenging task, which is often tackled using learning from demonstration datasets. The few garment folding datasets available nowadays to the robotics research community are either gathered from human demonstrations or generated through simulation. The former have the huge problem of perceiving human action and transferring it to the dynamic control of the robot, while the latter requires coding human motion into the simulator in open loop, resulting in far-from-realistic movements. In this article, we present a reduced but very accurate dataset of human cloth folding demonstrations. The dataset is collected through a novel virtual reality (VR) framework we propose, based on Unity’s 3D platform and the use of a HTC Vive Pro system. The framework is capable of simulating very realistic garments while allowing users to interact with them, in real time, through handheld controllers. By doing so, and thanks to the immersive experience, our framework gets rid of the gap between the human and robot perception-action loop, while simplifying data capture and resulting in more realistic sample, This work was developed in the context of the project CLOTHILDE (”CLOTH manIpulation Learning from DEmonstrations”) which has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement No. 741930) and is also supported by the BURG project PCI2019-103447 funded by MCIN/ AEI /10.13039/501100011033 and by the ”European Union”., Peer Reviewed, Postprint (published version)