Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Nov 26, 2025

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

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

Published on: December 15, 2023

Visual Robot Relocalization Based on Multi-Task CNN and Image-Similarity Strategy.

Tao Xie¹, Ke Wang¹, Ruifeng Li¹

¹State Key Laboratory of Robotics and System, Harbin Institute of Technology, 92 Xidazhi Street, Harbin 150006, China.

Sensors (Basel, Switzerland)

|December 9, 2020

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

From metabolic to pathologic complete response: case report of Neoadjuvant Lorlatinib in Stage IIIB ALK-positive NSCLC combined with uniportal VATS.

Journal of cardiothoracic surgery·2026

Same author

LUCAT1 promotes stemness in head and neck squamous cell carcinoma by sponging miR-128.

Frontiers in oncology·2026

Same author

AAV gene therapy for homozygous familial hypercholesterolemia: a phase 1 trial.

Nature medicine·2026

Same author

Mixed-methods study on GenAI Usage, dependence behaviors, and standardized application paths among Chinese medical students.

NPJ digital medicine·2026

Same author

Early supplemental feeding improves post-weaning growth restriction in lambs via the gastrointestinal-metabolic axis.

Applied and environmental microbiology·2026

Same author

Comparisons of thalamocortical functional connectivity in transgender women and cisgender individuals: associations with cognition in a Chinese cohort.

Dialogues in clinical neuroscience·2026

Same journal

RETRACTED: Zhang et al. A Novel Framework for Reconstruction and Imaging of Target Scattering Centers via Wide-Angle Incidence in Radar Networks. <i>Sensors</i> 2025, <i>25</i>, 6802.

Sensors (Basel, Switzerland)·2026

Same journal

Enhancing Unsupervised Multi-Source Domain Adaptation for Person Re-Identification via Mixture of Experts and Graph-Based Relation.

Sensors (Basel, Switzerland)·2026

Same journal

Development of an Instrumented Glove for Palmar Pressure Assessment in Kayakers.

Sensors (Basel, Switzerland)·2026

Same journal

Development and Experimental Validation of an Autonomous IoT-Based Monitoring System for Real-Time Water Quality Assessment in the Amazon River.

Sensors (Basel, Switzerland)·2026

Same journal

Semi-Supervised Adversarial Learning Framework for Controller Area Network Bus Intrusion Detection.

Sensors (Basel, Switzerland)·2026

Same journal

Smart Optimization Method for Safety Signs in Innovative Manufacturing Environments Integrating Industrial Field IoT Sensors and Knowledge Graphs.

Sensors (Basel, Switzerland)·2026

See all related articles

This study introduces a multi-task CNN for robot relocalization, improving accuracy and confidence by combining pose regression with scene recognition. A novel dual-level image-similarity strategy enhances robustness for real-time indoor and outdoor applications.

Area of Science:

Robotics
Computer Vision
Machine Learning

Background:

Traditional Convolutional Neural Networks (CNNs) for robot relocalization lack interpretability regarding prediction confidence.
Existing methods struggle with pose accuracy when visual differences between training and testing images are significant.

Purpose of the Study:

To develop a multi-task CNN for robot relocalization that simultaneously performs pose regression and scene recognition.
To enhance the robustness and accuracy of robot relocalization, especially under varying visual conditions.
To provide a confidence measure for pose estimations.

Main Methods:

Proposed a multi-task CNN integrating pose regression and scene recognition for enhanced relocalization.
Introduced a dual-level image-similarity strategy (DLISS) with initial and iteration levels (PSO-based image-block selection) to select visually similar testing images.

Keywords:

6D relocalization dual-level image-similarity strategy multi-task CNN scene recognition

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Related Experiment Videos

Last Updated: Nov 26, 2025

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

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

Published on: December 15, 2023

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Evaluated the method on the Microsoft 7Scenes and Cambridge Landmarks datasets.

Main Results:

Achieved approximately 0.33 m and 7.51° accuracy on 7Scenes, and 1.44 m and 4.83° on Cambridge Landmarks.
Reduced average positional error by 25% and angular error by 27.79% on 7Scenes compared to PoseNet.
Reduced average positional error by 40% and angular error by 28.55% on Cambridge Landmarks compared to PoseNet.

Conclusions:

The multi-task CNN effectively localizes using high-level features and is robust to out-of-scene images.
The DLISS strategy significantly improves relocalization accuracy by leveraging visually similar testing images.
The proposed method offers real-time performance (≤ 27 ms/frame) for both indoor and outdoor applications.