Improved object pose estimation via deep pre-touch sensing

Lancaster, Patrick; Yang, Boling; Smith, Joshua R.

doi:10.1109/iros.2017.8206061

Cited by 12 publications

(14 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The latter may result in imprecise estimation of object shape and pose with severe impacts on the grasping process. Although in-hand cameras are preferred to the more conventional head-mounted ones, there are still some limitations in the classical calibration procedures that result not so precise even though the distance between the sensor and the object is reduced, as demonstrated in [3].…”

Section: Introductionmentioning

confidence: 99%

“…Among the optical pre-touch sensors, Time-of-Flight (ToF) technology represents the solution to overcome the classical optical sensors issues. As shown in [3,12,13], time-offlight sensors do not need calibration and they are robust and accurate enough with a wide range of materials. In particular, Lancaster et al [3] successfully combine depth camera-based point cloud with pre-touch sensor information to improve the accuracy of the objects pose estimation, while Yang et al [12] used similar time-of-flight-based sensors to improve the accuracy of the grasping actions in sequential manipulation tasks during the Rubik's Cube solving.…”

Section: Introductionmentioning

confidence: 99%

“…As shown in [3,12,13], time-offlight sensors do not need calibration and they are robust and accurate enough with a wide range of materials. In particular, Lancaster et al [3] successfully combine depth camera-based point cloud with pre-touch sensor information to improve the accuracy of the objects pose estimation, while Yang et al [12] used similar time-of-flight-based sensors to improve the accuracy of the grasping actions in sequential manipulation tasks during the Rubik's Cube solving. Sasaki et al [13], more recently, successfully exploit ToF sensors to increase the robustness of the positioning of a robot hand during grasping task.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Proximity Sensor for Thin Wire Recognition and Manipulation

Cirillo¹,

Laudante²,

Pirozzi³

2021

Machines

View full text Add to dashboard Cite

In robotic grasping and manipulation, the knowledge of a precise object pose represents a key issue. The point acquires even more importance when the objects and, then, the grasping areas become smaller. This is the case of Deformable Linear Object manipulation application where the robot shall autonomously work with thin wires which pose and shape estimation could become difficult given the limited object size and possible occlusion conditions. In such applications, a vision-based system could not be enough to obtain accurate pose and shape estimation. In this work the authors propose a Time-of-Flight pre-touch sensor, integrated with a previously designed tactile sensor, for an accurate estimation of thin wire pose and shape. The paper presents the design and the characterization of the proposed sensor. Moreover, a specific object scanning and shape detection algorithm is presented. Experimental results support the proposed methodology, showing good performance. Hardware design and software applications are freely accessible to the reader.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Proximity Sensor for Thin Wire Recognition and Manipulation

Cirillo¹,

Laudante²,

Pirozzi³

2021

Machines

View full text Add to dashboard Cite

show abstract

“…The researchers realized grasping of tissue paper, a difficult task with only vision/depth and tactile sensor feedback. In [128], Lancaster et al study the use of deep learning to guide a proximitybased exploration strategy. It yields an improved object model as well as an improved estimate of the object's pose.…”

Section: Higher Complexity Methods and Behaviors (Towards Bt-ii)mentioning

confidence: 99%

“…The active perception principle has been studied in many robotics research papers, as discussed in a survey by Bohg et al [139]. Machine learning is essential in active perception for tasks such as object pose estimation or scene labeling, which can be based on proximity perception, e. g. [128]. Currently, a trend in active perception is the combined perception of vision and touch.…”

Section: B Cognitive Roboticsmentioning

confidence: 99%

Proximity Perception in Human-Centered Robotics: A Survey on Sensing Systems and Applications

et al. 2022

Self Cite

View full text Add to dashboard Cite

Wire Grasping by Using Proximity and Tactile Sensors

Cirillo

Laudante

Pirozzi

2022

2022 IEEE 5th International Conference on Industrial Cyber-Physical Systems (ICPS)

View full text Add to dashboard Cite

Nowadays robots have to be able to perform increasingly complex tasks. In grasping and manipulation, the knowledge of the environment and the pose of the target object are crucial for the correct execution of the task. Vision systems are widely used for environment and object perception, but they need to be finely calibrated to obtain high accuracy and they are not able to sense small objects like thin wires. Tactile sensors could be used to explore areas close to the target object, but this "blind" physical interaction is not always feasible. This paper proposes a strategy to use a proximity sensor mounted on the robot's end effector to obtain a pose estimation of the target object that, in this study, is represented by a thin electrical wire.

show abstract

Improved object pose estimation via deep pre-touch sensing

Cited by 12 publications

References 24 publications

Proximity Sensor for Thin Wire Recognition and Manipulation

Proximity Sensor for Thin Wire Recognition and Manipulation

Proximity Perception in Human-Centered Robotics: A Survey on Sensing Systems and Applications

Wire Grasping by Using Proximity and Tactile Sensors

Contact Info

Product

Resources

About