EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A multimodal tactile dataset for dynamic texture classification

Bruno Monteiro Rocha Lima, Venkata Naga Sai Siddhartha Danyamraju, Thiago Eustaquio Alves de Oliveira, Vinicius Prado da Fonseca

Data in Brief · 2023-10 · 50 · 109590

DOI: 10.1016/j.dib.2023.109590 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Data in Brief, 2023-10. Verfasst von Bruno Monteiro Rocha Lima und 3 weiteren Autor:innen. Beteiligte Einrichtung: School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, ON, Canada. Die Arbeit wurde bisher 17-mal zitiert (Crossref) und nennt 19 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reproducing human-like dexterous manipulation in robots requires identifying objects and textures. In unstructured settings, robots equipped with tactile sensors may detect textures by using touch-related characteristics. An extensive dataset of the physical interaction between a tactile-enable robotic probe is required to investigate and develop methods for categorizing textures. Therefore, this motivates us to compose a dataset from the signals of a bioinspired multimodal tactile sensing module while a robotic probe brings the module to dynamically contact 12 tactile textures under three exploratory velocities. This dataset contains pressure, acceleration, angular rate, and magnetic field variation signals from sensors embedded in the compliant structure of the sensing module. The pressure signals were sampled at 350 Hz, while the signals of the other sensors were sampled at 1500 Hz. Each texture was explored 100 times for each exploratory velocity, and each exploratory episode consisted of a sliding motion in the x and y directions tangential to the surface where the texture is placed. The total number of exploratory episodes in the dataset is 3600. The tactile texture dataset can be used for any project in the area of object recognition and robotic manipulation, and it is especially well suited for tactile texture reconstruction and recognition tasks. The dataset can also be used to study anisotropic textures and how robotic tactile exploration has to consider sliding motion directions.

Schlagwörter

Stichwörter der Autor:innen

Dynamic explorationMachine learningTactile sensorTexture classification

Autor:innen und Institutionen

  • School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, ON, Canada.
  • Department of Computer Science, Memorial University of Newfoundland, St. John's, NL, Canada.
  • Department of Computer Science, Lakehead University, Thunder Bay, ON, Canada.

Zitieren

APA

Monteiro Rocha Lima, B., Danyamraju, V. N. S. S., Alves de Oliveira, T. E., & Prado da Fonseca, V. (2023). A multimodal tactile dataset for dynamic texture classification. Data in Brief, 50, 109590. https://doi.org/10.1016/j.dib.2023.109590

Vancouver

Monteiro Rocha Lima B, Danyamraju VNSS, Alves de Oliveira TE, Prado da Fonseca V. A multimodal tactile dataset for dynamic texture classification. Data in Brief. 2023;50:109590. doi:10.1016/j.dib.2023.109590

BibTeX
@article{MonteiroRochaLima2023,
  author = {Monteiro Rocha Lima, Bruno and Danyamraju, Venkata Naga Sai Siddhartha and Alves de Oliveira, Thiago Eustaquio and Prado da Fonseca, Vinicius},
  title = {A multimodal tactile dataset for dynamic texture classification},
  journal = {Data in Brief},
  year = {2023},
  volume = {50},
  pages = {109590},
  doi = {10.1016/j.dib.2023.109590}
}
Kostenlos starten →