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

You might also read

Related Articles

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

Sort by
Same author

ORACLE: Object-Centric Autonomous Coverage Exploration Planner for Discrete Trunk Inspection Under Canopy.

Sensors (Basel, Switzerland)·2026
Same author

A Review of Multi-Sensor Fusion in Autonomous Driving.

Sensors (Basel, Switzerland)·2025
Same author

A Survey of Deep Learning-Driven 3D Object Detection: Sensor Modalities, Technical Architectures, and Applications.

Sensors (Basel, Switzerland)·2025
Same author

A Survey of the Multi-Sensor Fusion Object Detection Task in Autonomous Driving.

Sensors (Basel, Switzerland)·2025
Same author

A Review of Vision-Based Multi-Task Perception Research Methods for Autonomous Vehicles.

Sensors (Basel, Switzerland)·2025
Same author

Highway Safety with an Intelligent Headlight System for Improved Nighttime Driving.

Sensors (Basel, Switzerland)·2024

Related Experiment Video

Updated: Jul 10, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.8K

CompleteInst: An Efficient Instance Segmentation Network for Missed Detection Scene of Autonomous Driving.

Hai Wang1, Shilin Zhu1, Long Chen2

  • 1School of Automotive and Traffic Engineering, Jiangsu University, Zhenjiang 212013, China.

Sensors (Basel, Switzerland)
|November 25, 2023
PubMed
Summary
This summary is machine-generated.

CompleteInst enhances instance segmentation for autonomous driving by addressing missed detections. It improves recall on Cityscapes and COCO datasets using novel feature and instance level strategies.

Keywords:
autonomous drivingcomputer visioninstance segmentationmissed detection

More Related Videos

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.6K
A Method for 3D Reconstruction and Virtual Reality Analysis of Glial and Neuronal Cells
12:49

A Method for 3D Reconstruction and Virtual Reality Analysis of Glial and Neuronal Cells

Published on: September 28, 2019

12.8K

Related Experiment Videos

Last Updated: Jul 10, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.8K
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.6K
A Method for 3D Reconstruction and Virtual Reality Analysis of Glial and Neuronal Cells
12:49

A Method for 3D Reconstruction and Virtual Reality Analysis of Glial and Neuronal Cells

Published on: September 28, 2019

12.8K

Area of Science:

  • Computer Vision
  • Autonomous Driving Systems

Background:

  • Instance segmentation is crucial for autonomous driving, enabling both object distinction and pixel-level understanding.
  • Missed detections remain a significant challenge in current instance segmentation models.

Purpose of the Study:

  • To propose CompleteInst, a novel network architecture designed to mitigate missed detections in instance segmentation.
  • To improve the performance of autonomous driving systems through enhanced object recognition.

Main Methods:

  • Introduced Global Pyramid Networks (GPN) for collecting global information of missed instances at the feature level.
  • Developed a semantic branch to complete semantic features of missed instances.
  • Implemented a query-based optimal transport assignment (OTA-Query) for improved positive sample allocation at the instance level.
  • Designed parallel semantic and instance level branches compatible with QueryInst's supervision mechanism.

Main Results:

  • CompleteInst achieved recall rates of 56.7% on Cityscapes and 54.2% on COCO.
  • Demonstrated a performance improvement of 3.5% and 3.2% over the baseline, respectively.
  • The proposed parallel structure outperformed non-parallel alternatives.

Conclusions:

  • CompleteInst effectively reduces missed detections in instance segmentation tasks.
  • The novel feature and instance level strategies contribute to superior performance in autonomous driving applications.
  • The parallel design enhances efficiency and compatibility with existing frameworks.