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 Concept Videos

Detection of Black Holes01:10

Detection of Black Holes

2.3K
Although black holes were theoretically postulated in the 1920s, they remained outside the domain of observational astronomy until the 1970s.
Their closest cousins are neutron stars, which are composed almost entirely of neutrons packed against each other, making them extremely dense. A neutron star has the same mass as the Sun but its diameter is only a few kilometers. Therefore, the escape velocity from their surface is close to the speed of light.
Not until the 1960s, when the first neutron...
2.3K
  1. Home
  3. Research Domains
  5. Engineering
  7. Environmental Engineering
  9. Air Pollution Modelling And Control
  11. Mine-dw-fusion: Bev Multiscale-enhanced Fusion Object-detection Model For Underground Coal Mine Based On Dynamic Weight Adjustment

Mine-DW-Fusion: BEV Multiscale-Enhanced Fusion Object-Detection Model for Underground Coal Mine Based on Dynamic Weight Adjustment

Wanzi Yan1, Yidong Zhang2, Minti Xue3

  • 1School of Mines, China University of Mining and Technology, Xuzhou 221116, China.

Sensors (Basel, Switzerland)
|August 28, 2025

Related Experiment 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

635
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

6.9K

View abstract on PubMed

Summary
This summary is machine-generated.

This study introduces a novel perception model for autonomous mining vehicles, enhancing safety in challenging underground conditions. The model effectively fuses sensor data, even with failures, improving object detection in low light and dust.

Area of Science:

  • Robotics and Autonomous Systems
  • Computer Vision
  • Geoscience Engineering

Background:

  • Autonomous driving in underground coal mines faces perception challenges due to dust, poor lighting, and sensor abnormalities.
  • Existing methods lack effective reliability evaluation for multimodal sensor data and robust fusion techniques for sensor failures.
  • A dedicated multimodal dataset for underground coal mine environments is needed for model development and validation.

Purpose of the Study:

  • To develop an advanced multimodal fusion perception model for autonomous haulage vehicles in underground coal mines.
  • To address the limitations in evaluating data reliability and handling sensor failures in complex mining environments.
  • To create a comprehensive dataset for training and testing perception models in real-world underground coal mine scenarios.
Keywords:
autonomous drivingauxiliary transportation vehiclesbird’s-eye viewenvironmental perception

Related Experiment 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

635
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

6.9K

Main Methods:

  • Proposed a Bird's-Eye View (BEV) multiscale-enhanced fusion perception model with dynamic weight adjustment.
  • Implemented a Mixture of Experts-Fuzzy Logic Inference Module (MoE-FLIM) for modality data weighting based on BEV features.
  • Introduced a Pyramid Multiscale Feature Enhancement and Fusion Module (PMS-FFEM) to handle sensor data abnormalities.
  • Constructed a multimodal dataset specifically for underground coal mine environments.

Main Results:

  • The proposed model achieves high accuracy and stability in object detection within underground coal mine settings.
  • Demonstrated robust perception performance in challenging conditions, including low light and dust fog.
  • The dynamic weight adjustment and multiscale fusion effectively mitigate issues arising from sensor data abnormalities.

Conclusions:

  • The developed BEV multiscale-enhanced fusion perception model significantly improves autonomous driving capabilities in underground coal mines.
  • The MoE-FLIM and PMS-FFEM modules provide effective solutions for data reliability assessment and sensor failure handling.
  • The created dataset supports the advancement of perception technologies for safer and more efficient mining operations.
multimodal information fusion