Memory Unscented Particle Filter for 6-DOF Tactile Localization
This paper addresses 6-DOF (degree-of-freedom) tactile localization, i.e. the pose estimation of tridimensional objects given tactile measurements. This estimation problem is fundamental for the operation of autonomous robots that are often required to manipulate and grasp objects whose pose is a-pr...
Đã lưu trong:
| Những tác giả chính: | , , , , |
|---|---|
| Định dạng: | Bài viết |
| Được phát hành: |
2016
|
| Những chủ đề: | |
| Truy cập trực tuyến: | https://arxiv.org/abs/1607.02757 http://arxiv.org/abs/1607.02757 |
| Các nhãn: |
Không có thẻ, Là người đầu tiên thẻ bản ghi này!
|
| _version_ | 1879177603890282496 |
|---|---|
| author | Vezzani, Giulia Pattacini, Ugo Battistelli, Giorgio Chisci, Luigi Natale, Lorenzo |
| author_browse | Battistelli, Giorgio Chisci, Luigi Natale, Lorenzo Pattacini, Ugo Vezzani, Giulia |
| author_facet | Vezzani, Giulia Pattacini, Ugo Battistelli, Giorgio Chisci, Luigi Natale, Lorenzo |
| author_sort | Vezzani, Giulia |
| building | arXiv |
| collection | Acceso abierto |
| description | This paper addresses 6-DOF (degree-of-freedom) tactile localization, i.e. the pose estimation of tridimensional objects given tactile measurements. This estimation problem is fundamental for the operation of autonomous robots that are often required to manipulate and grasp objects whose pose is a-priori unknown. The nature of tactile measurements, the strict time requirements for real-time operation and the multimodality of the involved probability distributions pose remarkable challenges and call for advanced nonlinear filtering techniques. Following a Bayesian approach, this paper proposes a novel and effective algorithm, named Memory Unscented Particle Filter (MUPF), which solves the 6-DOF localization problem recursively in real-time by only exploiting contact point measurements. MUPF combines a modified particle filter that incorporates a sliding memory of past measurements to better handle multimodal distributions, along with the unscented Kalman filter that moves the particles towards regions of the search space that are more likely with the measurements. The performance of the proposed MUPF algorithm has been assessed both in simulation and on a real robotic system equipped with tactile sensors (i.e., the iCub humanoid robot). The experiments show that the algorithm provides accurate and reliable localization even with a low number of particles and, hence, is compatible with real-time requirements. |
| doi_str_mv | 10.1109/TRO.2017.2707092 |
| first_indexed | 2026-10-16T01:03:25Z |
| format | Article |
| fullrecord | <oai_dc:dc xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ https://www.openarchives.org/OAI/2.0/oai_dc.xsd"><identifier>arxiv-1607.02757</identifier><datestamp>2025-09-16</datestamp>
<dc:title>Memory Unscented Particle Filter for 6-DOF Tactile Localization</dc:title>
<dc:creator>Vezzani, Giulia</dc:creator>
<dc:creator>Pattacini, Ugo</dc:creator>
<dc:creator>Battistelli, Giorgio</dc:creator>
<dc:creator>Chisci, Luigi</dc:creator>
<dc:creator>Natale, Lorenzo</dc:creator>
<dc:subject>Robotics</dc:subject>
<dc:description>This paper addresses 6-DOF (degree-of-freedom) tactile localization, i.e. the pose estimation of tridimensional objects given tactile measurements. This estimation problem is fundamental for the operation of autonomous robots that are often required to manipulate and grasp objects whose pose is a-priori unknown. The nature of tactile measurements, the strict time requirements for real-time operation and the multimodality of the involved probability distributions pose remarkable challenges and call for advanced nonlinear filtering techniques. Following a Bayesian approach, this paper proposes a novel and effective algorithm, named Memory Unscented Particle Filter (MUPF), which solves the 6-DOF localization problem recursively in real-time by only exploiting contact point measurements. MUPF combines a modified particle filter that incorporates a sliding memory of past measurements to better handle multimodal distributions, along with the unscented Kalman filter that moves the particles towards regions of the search space that are more likely with the measurements. The performance of the proposed MUPF algorithm has been assessed both in simulation and on a real robotic system equipped with tactile sensors (i.e., the iCub humanoid robot). The experiments show that the algorithm provides accurate and reliable localization even with a low number of particles and, hence, is compatible with real-time requirements. </dc:description>
<dc:date>2016-07-10</dc:date>
<dc:date>2016-11-10</dc:date>
<dc:type>text</dc:type>
<dc:identifier>https://arxiv.org/abs/1607.02757</dc:identifier>
<dc:identifier>http://arxiv.org/abs/1607.02757</dc:identifier>
<dc:identifier>IEEE Transactions on Robotics, Volume 33, Issue 5, October 2017</dc:identifier>
<dc:identifier>doi:10.1109/TRO.2017.2707092</dc:identifier>
</oai_dc:dc> |
| id | arxiv-1607.02757 |
| institution | arXiv |
| last_indexed | 2026-10-16T01:03:25Z |
| publishDate | 2016 |
| publishDateRange | 2016 |
| publishDateSort | 2016 |
| record_format | oai_dc |
| spelling | arxiv-1607.027572025-09-16 Memory Unscented Particle Filter for 6-DOF Tactile Localization Vezzani, Giulia Pattacini, Ugo Battistelli, Giorgio Chisci, Luigi Natale, Lorenzo Robotics This paper addresses 6-DOF (degree-of-freedom) tactile localization, i.e. the pose estimation of tridimensional objects given tactile measurements. This estimation problem is fundamental for the operation of autonomous robots that are often required to manipulate and grasp objects whose pose is a-priori unknown. The nature of tactile measurements, the strict time requirements for real-time operation and the multimodality of the involved probability distributions pose remarkable challenges and call for advanced nonlinear filtering techniques. Following a Bayesian approach, this paper proposes a novel and effective algorithm, named Memory Unscented Particle Filter (MUPF), which solves the 6-DOF localization problem recursively in real-time by only exploiting contact point measurements. MUPF combines a modified particle filter that incorporates a sliding memory of past measurements to better handle multimodal distributions, along with the unscented Kalman filter that moves the particles towards regions of the search space that are more likely with the measurements. The performance of the proposed MUPF algorithm has been assessed both in simulation and on a real robotic system equipped with tactile sensors (i.e., the iCub humanoid robot). The experiments show that the algorithm provides accurate and reliable localization even with a low number of particles and, hence, is compatible with real-time requirements. 2016-07-10 2016-11-10 text https://arxiv.org/abs/1607.02757 http://arxiv.org/abs/1607.02757 IEEE Transactions on Robotics, Volume 33, Issue 5, October 2017 doi:10.1109/TRO.2017.2707092 |
| spellingShingle | Robotics Vezzani, Giulia Pattacini, Ugo Battistelli, Giorgio Chisci, Luigi Natale, Lorenzo Memory Unscented Particle Filter for 6-DOF Tactile Localization |
| title | Memory Unscented Particle Filter for 6-DOF Tactile Localization |
| title_full | Memory Unscented Particle Filter for 6-DOF Tactile Localization |
| title_fullStr | Memory Unscented Particle Filter for 6-DOF Tactile Localization |
| title_full_unstemmed | Memory Unscented Particle Filter for 6-DOF Tactile Localization |
| title_short | Memory Unscented Particle Filter for 6-DOF Tactile Localization |
| title_sort | memory unscented particle filter for 6 dof tactile localization |
| topic | Robotics |
| url | https://arxiv.org/abs/1607.02757 http://arxiv.org/abs/1607.02757 |
| work_keys_str_mv | AT vezzanigiulia memoryunscentedparticlefilterfor6doftactilelocalization AT pattaciniugo memoryunscentedparticlefilterfor6doftactilelocalization AT battistelligiorgio memoryunscentedparticlefilterfor6doftactilelocalization AT chisciluigi memoryunscentedparticlefilterfor6doftactilelocalization AT natalelorenzo memoryunscentedparticlefilterfor6doftactilelocalization |