An End-to-End Grasping Stability Prediction Network for Multiple Sensors
As we all know, the output of the tactile sensing array on the gripper can be used to predict grasping stability. Some methods utilize traditional tactile features to make the decision and some advanced methods use machine learning or deep learning ways to build a prediction model. While these metho...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-03-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/10/6/1997 |
id |
doaj-3968b150041f4a1db7dc5d072a93cd9a |
---|---|
record_format |
Article |
spelling |
doaj-3968b150041f4a1db7dc5d072a93cd9a2020-11-25T02:28:41ZengMDPI AGApplied Sciences2076-34172020-03-01106199710.3390/app10061997app10061997An End-to-End Grasping Stability Prediction Network for Multiple SensorsXin Shu0Chang Liu1Tong Li2National Institute of Standards and Tec State Key Laboratory Transducer, Institute of Electronics, Chinese Academy of Sciences, Beijing 100190, ChinaNational Institute of Standards and Tec State Key Laboratory Transducer, Institute of Electronics, Chinese Academy of Sciences, Beijing 100190, ChinaNational Institute of Standards and Tec State Key Laboratory Transducer, Institute of Electronics, Chinese Academy of Sciences, Beijing 100190, ChinaAs we all know, the output of the tactile sensing array on the gripper can be used to predict grasping stability. Some methods utilize traditional tactile features to make the decision and some advanced methods use machine learning or deep learning ways to build a prediction model. While these methods are all limited to the specific sensing array and have two common disadvantages. On the one hand, these models cannot perform well on different sensors. On the other hand, they do not have the ability of inferencing on multiple sensors in an end-to-end manner. Thus, we aim to find the internal relationships among different sensors and inference the grasping stability of multiple sensors in an end-to-end way. In this paper, we propose the MM-CNN (mask multi-head convolutional neural network), which can be utilized to predict the grasping stability on the output of multiple sensors with the weight sharing mechanism. We train this model and evaluate it on our own collected datasets. This model achieves 99.49% and 94.25% prediction accuracy on two different sensing arrays, separately. In addition, we show that our proposed structure is also available for other CNN backbones and can be easily integrated.https://www.mdpi.com/2076-3417/10/6/1997robotic graspingtactile perceptionintelligent manipulationstability prediction |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Xin Shu Chang Liu Tong Li |
spellingShingle |
Xin Shu Chang Liu Tong Li An End-to-End Grasping Stability Prediction Network for Multiple Sensors Applied Sciences robotic grasping tactile perception intelligent manipulation stability prediction |
author_facet |
Xin Shu Chang Liu Tong Li |
author_sort |
Xin Shu |
title |
An End-to-End Grasping Stability Prediction Network for Multiple Sensors |
title_short |
An End-to-End Grasping Stability Prediction Network for Multiple Sensors |
title_full |
An End-to-End Grasping Stability Prediction Network for Multiple Sensors |
title_fullStr |
An End-to-End Grasping Stability Prediction Network for Multiple Sensors |
title_full_unstemmed |
An End-to-End Grasping Stability Prediction Network for Multiple Sensors |
title_sort |
end-to-end grasping stability prediction network for multiple sensors |
publisher |
MDPI AG |
series |
Applied Sciences |
issn |
2076-3417 |
publishDate |
2020-03-01 |
description |
As we all know, the output of the tactile sensing array on the gripper can be used to predict grasping stability. Some methods utilize traditional tactile features to make the decision and some advanced methods use machine learning or deep learning ways to build a prediction model. While these methods are all limited to the specific sensing array and have two common disadvantages. On the one hand, these models cannot perform well on different sensors. On the other hand, they do not have the ability of inferencing on multiple sensors in an end-to-end manner. Thus, we aim to find the internal relationships among different sensors and inference the grasping stability of multiple sensors in an end-to-end way. In this paper, we propose the MM-CNN (mask multi-head convolutional neural network), which can be utilized to predict the grasping stability on the output of multiple sensors with the weight sharing mechanism. We train this model and evaluate it on our own collected datasets. This model achieves 99.49% and 94.25% prediction accuracy on two different sensing arrays, separately. In addition, we show that our proposed structure is also available for other CNN backbones and can be easily integrated. |
topic |
robotic grasping tactile perception intelligent manipulation stability prediction |
url |
https://www.mdpi.com/2076-3417/10/6/1997 |
work_keys_str_mv |
AT xinshu anendtoendgraspingstabilitypredictionnetworkformultiplesensors AT changliu anendtoendgraspingstabilitypredictionnetworkformultiplesensors AT tongli anendtoendgraspingstabilitypredictionnetworkformultiplesensors AT xinshu endtoendgraspingstabilitypredictionnetworkformultiplesensors AT changliu endtoendgraspingstabilitypredictionnetworkformultiplesensors AT tongli endtoendgraspingstabilitypredictionnetworkformultiplesensors |
_version_ |
1724837135180103680 |