Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jul 30, 2025

Investigating Object Representations in the Macaque Dorsal Visual Stream Using Single-unit Recordings
07:08

Investigating Object Representations in the Macaque Dorsal Visual Stream Using Single-unit Recordings

Published on: August 1, 2018

8.4K

Gradient adaptive sampling and multiple temporal scale 3D CNNs for tactile object recognition.

Xiaoliang Qian1, Jia Meng1, Wei Wang1

  • 1College of Electrical and Information Engineering, Zhengzhou University of Light Industry, Zhengzhou, China.

Frontiers in Neurorobotics
|May 14, 2023
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Comparison of posterior scleral reinforcement surgery and repeated low-level red-light therapy in controlling high and super high myopia in Chinese children.

International journal of ophthalmology·2026
Same author

VeloRM: disentangling pre- and post-splicing RNA modification dynamics at single-cell resolution.

Nucleic acids research·2026
Same author

DirectASRM: Uncovering allele-specific post-transcriptional RNA modifications through direct RNA sequencing.

Bioinformatics (Oxford, England)·2026
Same author

Unraveling the mechanisms of micro-voltage-driven fungal remediation for simultaneous removal of microplastics, antibiotics, and heavy metals.

Journal of hazardous materials·2026
Same author

Application of indocyanine green fluorescence imaging in cortical-sparing adrenalectomy: a preliminary study with image quantification techniques.

Surgical endoscopy·2026
Same author

Sequence determinant and functional relevance of 8-oxoguanine RNA modification unveiled from foundation-model-based predictor.

Molecular therapy. Nucleic acids·2026
Same journal

DSPE-ViT: a lightweight vision transformer with dynamic sparse positional encoding for dense small object detection in UAV imagery.

Frontiers in neurorobotics·2026
Same journal

ST-HONet: Spatio-Temporal Hierarchical Network for long-horizon bimanual visuomotor imitation.

Frontiers in neurorobotics·2026
Same journal

ST-HADP: Spatio-Temporal hierarchical attention diffusion policy for long-horizon generalizable bimanual visuomotor imitation.

Frontiers in neurorobotics·2026
Same journal

EQISP: efficient quantized image signal processing with multi-scale pyramid fusion for resource constrained embodied perception.

Frontiers in neurorobotics·2026
Same journal

Research on embodied agent multimodal perception and real-time path planning algorithms for complex unstructured environments.

Frontiers in neurorobotics·2026
Same journal

NL-YOLOv5: a model with a larger receptive field and the ability to globally acquire features.

Frontiers in neurorobotics·2026
See all related articles

This study introduces gradient adaptive sampling (GAS) and multiple temporal scale 3D convolutional neural networks (MTS-3DCNNs) to improve tactile object recognition (TOR) in robots. These methods enhance data efficiency and generalization across different grasping speeds.

Area of Science:

  • Robotics
  • Artificial Intelligence
  • Computer Vision

Background:

  • Tactile object recognition (TOR) is crucial for robotic perception.
  • Existing TOR methods struggle with data redundancy or information loss due to uniform sampling.
  • Current models lack generalization for varying grasping speeds due to single time-scale processing.

Purpose of the Study:

  • To develop a more efficient and generalizable TOR method.
  • To address the limitations of uniform sampling and single time-scale processing in TOR.
  • To improve the accuracy and robustness of robotic tactile perception.

Main Methods:

  • Proposed a novel gradient adaptive sampling (GAS) strategy to prioritize important tactile data.
  • Introduced a multiple temporal scale 3D convolutional neural networks (MTS-3DCNNs) model for enhanced feature extraction.
Keywords:
3D convolutional neural networksMR3D-18 networkgradient adaptive samplingmultiple temporal scaletactile object recognition

More Related Videos

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

591
Adaptation of a Haptic Robot in a 3T fMRI
08:16

Adaptation of a Haptic Robot in a 3T fMRI

Published on: October 4, 2011

9.8K

Related Experiment Videos

Last Updated: Jul 30, 2025

Investigating Object Representations in the Macaque Dorsal Visual Stream Using Single-unit Recordings
07:08

Investigating Object Representations in the Macaque Dorsal Visual Stream Using Single-unit Recordings

Published on: August 1, 2018

8.4K
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

591
Adaptation of a Haptic Robot in a 3T fMRI
08:16

Adaptation of a Haptic Robot in a 3T fMRI

Published on: October 4, 2011

9.8K
  • Modified ResNet3D-18 into a lightweight MR3D-18 network to prevent overfitting with smaller tactile datasets.

    • Main Results:

    • Ablation studies confirmed the effectiveness of the GAS strategy, MTS-3DCNNs, and MR3D-18 networks.
    • The proposed methods significantly improved data efficiency by adaptively selecting key tactile frames.
    • The MTS-3DCNNs model demonstrated superior generalization capabilities for objects grasped at different speeds.

    Conclusions:

    • The developed GAS and MTS-3DCNNs approach represents a significant advancement in tactile object recognition.
    • The MR3D-18 network provides an efficient solution for processing tactile data while mitigating overfitting.
    • The proposed methods achieve state-of-the-art performance on benchmark datasets, enhancing robotic tactile perception.