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: Jun 19, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

Transformer-CNN Hybrid Framework for Pavement Pothole Segmentation.

Tianjie Zhang1,2, Zhen Liu3, Bingyan Cui1

  • 1Center for Advanced Infrastructure and Transportation, Rutgers University, Piscataway, NJ 08854, USA.

Sensors (Basel, Switzerland)
|November 13, 2025
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

Investigation of damage evolution in asphalt concrete under coupled ultraviolet aging and freeze-thaw cycles.

Scientific reports·2026
Same author

How urban environments structure running behaviour in Beijing during winter, spring, and summer 2024: spatiotemporal patterns and configuration-specific interactions from running trajectory data.

Archives of public health = Archives belges de sante publique·2026
Same author

Identification and Evaluation of Volatile Compounds from Solanaceae Plants Attractive to <i>Tuta absoluta</i> (Lepidoptera: Gelechiidae).

Journal of agricultural and food chemistry·2026
Same author

PRC1 promotes MAFLD progression by regulating glycolysis/lactylation axis and forming a positive feedback loop.

The Journal of nutritional biochemistry·2026
Same author

Uniform efficacy of SGLT2 inhibitors across the ejection fraction spectrum and in high-risk patients with HFpEF: a prespecified pooled analysis.

Frontiers in cardiovascular medicine·2026
Same author

Betulin Protects Against Cardiac Hypertrophy by Improving AMPK/Nrf2-Dependent Mitochondrial Function.

Phytotherapy research : PTR·2026

This study introduces PoFormer, a hybrid deep learning model for accurate pothole detection. PoFormer enhances pavement safety and infrastructure maintenance by improving automated defect segmentation.

Area of Science:

  • Civil Engineering
  • Computer Science
  • Artificial Intelligence

Background:

  • Pavement surface defects like potholes present significant safety hazards and accelerate infrastructure decay.
  • Automated detection of these defects necessitates advanced sensing and deep learning models.

Purpose of the Study:

  • To propose PoFormer, a Transformer-CNN hybrid framework for precise pavement pothole segmentation.
  • To evaluate PoFormer's performance against state-of-the-art models using a diverse dataset.

Main Methods:

  • Developed a Transformer-CNN hybrid framework (PoFormer) integrating global and local feature extraction.
  • Created a comprehensive dataset using open-source images and high-resolution field data from a multi-sensor inspection vehicle.
  • Employed millimeter-level resolution 3D surface imaging with infrared/laser-assisted lighting for robust data acquisition.
Keywords:
CNNdeep learningimage segmentationpotholetransformer

Related Experiment Videos

Last Updated: Jun 19, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

Main Results:

  • PoFormer achieved a mean Intersection over Union (IoU) of 77.23%.
  • The model obtained a mean pixel accuracy of 84.48%.
  • PoFormer demonstrated superior segmentation accuracy compared to existing CNN-based models.

Conclusions:

  • The proposed PoFormer framework offers superior performance in pavement pothole segmentation.
  • Integrating multi-sensor 3D imaging with hybrid neural networks advances intelligent pavement condition monitoring.
  • This technology holds significant potential for automated infrastructure maintenance and road safety improvement.