2021 IEEE International Conference on Robotics and Automation (ICRA) 2021
DOI: 10.1109/icra48506.2021.9561122
|View full text |Cite
|
Sign up to set email alerts
|

Simulation of Vision-based Tactile Sensors using Physics based Rendering

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
24
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
3
3
1

Relationship

1
6

Authors

Journals

citations
Cited by 31 publications
(24 citation statements)
references
References 21 publications
0
24
0
Order By: Relevance
“…We first introduce the computation cost in Subsection V-A. To evaluate our approach, Subsection V-B compares the pseudo tactile images from [1], [19] and Tacchi. We also show images with different particle numbers to explore the effect of this parameter.…”
Section: Resultsmentioning
confidence: 99%
See 2 more Smart Citations
“…We first introduce the computation cost in Subsection V-A. To evaluate our approach, Subsection V-B compares the pseudo tactile images from [1], [19] and Tacchi. We also show images with different particle numbers to explore the effect of this parameter.…”
Section: Resultsmentioning
confidence: 99%
“…Experimental results show that Tacchi can be connected with robotics simulators like Gazebo and MuJoCo. Our method produces higher quality tactile images than the existing methods [1], [19] in terms of both tactile image quality and Sim2Real transfer learning. In addition, Tacchi can add random defects to simulate textures on real objects and improve the Sim2Real performance.…”
mentioning
confidence: 87%
See 1 more Smart Citation
“…Existing work [23], [24], [25], [26] build the mechanics models to simulate the soft body deformation for tactile sensors. For vision-based tactile sensors like GelSight [27], optical simulation [9], [10], [8], [11] is also essential as it is used to measure shapes of objects in contact. Several simto-real robot perception work [5], [16], [28] has revealed the accuracy and sufficiency of tactile optical simulation, and we leverage the existing tactile sensing simulation in robot simulator to make better use of them.…”
Section: B Simulation Of Tactile Sensorsmentioning
confidence: 99%
“…When solving these problems with data-driven approaches, simulating accurate contact dynamics and dense visionbased tactile readings can provide efficient data collection. There exists multiple vision-based tactile simulation [8], [9], [10], [11] which takes contact shapes as inputs and simulate the tactile images, and [12], [13] integrate the tactile simulation into robot dynamics simulation engines. However, these simulators are limited for dexterous manipulation tasks which also requires accurate contact dynamics simulation to handle contact forces, friction and the converting from contact physics to contact shapes of tactile sensors.…”
Section: Introductionmentioning
confidence: 99%