Robotic pick-and-place of novel objects in clutter with multi-affordance grasping and cross-domain image matching
This article presents a robotic pick-and-place system that is capable of grasping and recognizing both known and novel objects in cluttered environments. The key new feature of the system is that it handles a wide range of object categories without needing any task-specific training data for novel o...
| Main Authors: | , , , , , , , , , , , , , , , , , , , , |
|---|---|
| 其他作者: | |
| 格式: | 文件 |
| 语言: | English |
| 出版: |
SAGE Publications
2021
|
| 在线阅读: | https://hdl.handle.net/1721.1/130311 |
| _version_ | 1826214526473732096 |
|---|---|
| author | Zeng, Andy Song, Shuran Yu, Kuan-Ting Donlon, Elliott S Hogan, Francois R. Bauza Villalonga, Maria Ma, Daolin Taylor, Orion Thomas Liu, Melody Romo, Eudald Fazeli, Nima Alet, Ferran Chavan Dafle, Nikhil Narsingh Holladay, Rachel Morona, Isabella Nair, Prem Qu Green, Druck Taylor, Ian Liu, Weber Funkhouser, Thomas Rodriguez, Alberto |
| author2 | Massachusetts Institute of Technology. Department of Mechanical Engineering |
| author_facet | Massachusetts Institute of Technology. Department of Mechanical Engineering Zeng, Andy Song, Shuran Yu, Kuan-Ting Donlon, Elliott S Hogan, Francois R. Bauza Villalonga, Maria Ma, Daolin Taylor, Orion Thomas Liu, Melody Romo, Eudald Fazeli, Nima Alet, Ferran Chavan Dafle, Nikhil Narsingh Holladay, Rachel Morona, Isabella Nair, Prem Qu Green, Druck Taylor, Ian Liu, Weber Funkhouser, Thomas Rodriguez, Alberto |
| author_sort | Zeng, Andy |
| collection | MIT |
| description | This article presents a robotic pick-and-place system that is capable of grasping and recognizing both known and novel objects in cluttered environments. The key new feature of the system is that it handles a wide range of object categories without needing any task-specific training data for novel objects. To achieve this, it first uses an object-agnostic grasping framework to map from visual observations to actions: inferring dense pixel-wise probability maps of the affordances for four different grasping primitive actions. It then executes the action with the highest affordance and recognizes picked objects with a cross-domain image classification framework that matches observed images to product images. Since product images are readily available for a wide range of objects (e.g., from the web), the system works out-of-the-box for novel objects without requiring any additional data collection or re-training. Exhaustive experimental results demonstrate that our multi-affordance grasping achieves high success rates for a wide variety of objects in clutter, and our recognition algorithm achieves high accuracy for both known and novel grasped objects. The approach was part of the MIT–Princeton Team system that took first place in the stowing task at the 2017 Amazon Robotics Challenge. All code, datasets, and pre-trained models are available online at http://arc.cs.princeton.edu/ |
| first_indexed | 2024-09-23T16:06:50Z |
| format | Article |
| id | mit-1721.1/130311 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T16:06:50Z |
| publishDate | 2021 |
| publisher | SAGE Publications |
| record_format | dspace |
| spelling | mit-1721.1/1303112022-09-29T18:19:19Z Robotic pick-and-place of novel objects in clutter with multi-affordance grasping and cross-domain image matching Zeng, Andy Song, Shuran Yu, Kuan-Ting Donlon, Elliott S Hogan, Francois R. Bauza Villalonga, Maria Ma, Daolin Taylor, Orion Thomas Liu, Melody Romo, Eudald Fazeli, Nima Alet, Ferran Chavan Dafle, Nikhil Narsingh Holladay, Rachel Morona, Isabella Nair, Prem Qu Green, Druck Taylor, Ian Liu, Weber Funkhouser, Thomas Rodriguez, Alberto Massachusetts Institute of Technology. Department of Mechanical Engineering This article presents a robotic pick-and-place system that is capable of grasping and recognizing both known and novel objects in cluttered environments. The key new feature of the system is that it handles a wide range of object categories without needing any task-specific training data for novel objects. To achieve this, it first uses an object-agnostic grasping framework to map from visual observations to actions: inferring dense pixel-wise probability maps of the affordances for four different grasping primitive actions. It then executes the action with the highest affordance and recognizes picked objects with a cross-domain image classification framework that matches observed images to product images. Since product images are readily available for a wide range of objects (e.g., from the web), the system works out-of-the-box for novel objects without requiring any additional data collection or re-training. Exhaustive experimental results demonstrate that our multi-affordance grasping achieves high success rates for a wide variety of objects in clutter, and our recognition algorithm achieves high accuracy for both known and novel grasped objects. The approach was part of the MIT–Princeton Team system that took first place in the stowing task at the 2017 Amazon Robotics Challenge. All code, datasets, and pre-trained models are available online at http://arc.cs.princeton.edu/ NSF (Grants IIS-1251217, VEC 1539014/1539099) 2021-03-31T19:02:14Z 2021-03-31T19:02:14Z 2019-08 2020-08-03T13:55:36Z Article http://purl.org/eprint/type/JournalArticle 0278-3649 1741-3176 https://hdl.handle.net/1721.1/130311 Zeng, Andy et al. "Robotic pick-and-place of novel objects in clutter with multi-affordance grasping and cross-domain image matching." International Journal of Robotics Research (August 2019): 1-16. en 10.1177/0278364919868017 International Journal of Robotics Research Creative Commons Attribution-NonCommercial-NoDerivs License http://creativecommons.org/licenses/by-nc-nd/4.0/ application/pdf SAGE Publications Sage |
| spellingShingle | Zeng, Andy Song, Shuran Yu, Kuan-Ting Donlon, Elliott S Hogan, Francois R. Bauza Villalonga, Maria Ma, Daolin Taylor, Orion Thomas Liu, Melody Romo, Eudald Fazeli, Nima Alet, Ferran Chavan Dafle, Nikhil Narsingh Holladay, Rachel Morona, Isabella Nair, Prem Qu Green, Druck Taylor, Ian Liu, Weber Funkhouser, Thomas Rodriguez, Alberto Robotic pick-and-place of novel objects in clutter with multi-affordance grasping and cross-domain image matching |
| title | Robotic pick-and-place of novel objects in clutter with multi-affordance grasping and cross-domain image matching |
| title_full | Robotic pick-and-place of novel objects in clutter with multi-affordance grasping and cross-domain image matching |
| title_fullStr | Robotic pick-and-place of novel objects in clutter with multi-affordance grasping and cross-domain image matching |
| title_full_unstemmed | Robotic pick-and-place of novel objects in clutter with multi-affordance grasping and cross-domain image matching |
| title_short | Robotic pick-and-place of novel objects in clutter with multi-affordance grasping and cross-domain image matching |
| title_sort | robotic pick and place of novel objects in clutter with multi affordance grasping and cross domain image matching |
| url | https://hdl.handle.net/1721.1/130311 |
| work_keys_str_mv | AT zengandy roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT songshuran roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT yukuanting roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT donlonelliotts roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT hoganfrancoisr roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT bauzavillalongamaria roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT madaolin roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT taylororionthomas roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT liumelody roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT romoeudald roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT fazelinima roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT aletferran roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT chavandaflenikhilnarsingh roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT holladayrachel roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT moronaisabella roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT nairpremqu roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT greendruck roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT taylorian roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT liuweber roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT funkhouserthomas roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching AT rodriguezalberto roboticpickandplaceofnovelobjectsinclutterwithmultiaffordancegraspingandcrossdomainimagematching |