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

Integrative transcriptomics and scpagwas analysis predicts immune-related markers of preeclampsia.

Journal of reproductive immunology·2026
Same author

Eerdun Wurile, a Mongolian medicine, alleviates neuronal damage and improves neurological dysfunction after cerebral hemorrhage by activating the PI3K/AKT signaling pathway.

Chinese medicine·2026
Same author

Mass Spectrometry-Based Lipidomics in Coffee: Linking Lipid Transformation to Flavor Formation and Quality Control.

Foods (Basel, Switzerland)·2026
Same author

Publisher Correction: In situ nanocrystal confinement for efficient blue perovskite LEDs.

Nature·2026
Same author

GLP-1 Receptor Agonists in Neurological Disorders: From Mechanisms to Clinical Translation.

Drug design, development and therapy·2026
Same author

Study on the Functional, Physiological, and Sensory Properties of Coffee Leaf Dark Tea Processed Using the Pile Fermentation Method of Pu-Erh Tea.

Foods (Basel, Switzerland)·2026

Related Experiment Video

Updated: Apr 9, 2026

Author Spotlight: Insights into the Analysis of Human Interaction with 3D Virtual Objects
06:36

Author Spotlight: Insights into the Analysis of Human Interaction with 3D Virtual Objects

Published on: October 18, 2024

1.5K

LiteMapNet: Lightweight Online High-Definition Mapping With Attention Head Pruning.

Shuangtong Liu1,2,3, Tao Liu1,2,3, Wenning Wang4

  • 1Faculty of Geomatics, Lanzhou Jiaotong University, Lanzhou, China.

Annals of the New York Academy of Sciences
|April 7, 2026
PubMed
Summary

LiteMapNet offers a lightweight solution for real-time high-definition (HD) map construction in autonomous driving. This transformer-based framework significantly reduces computational demands, enabling efficient deployment on resource-limited vehicles.

Keywords:
Fisher information matrixTransformerattention‐head pruningonline HD map constructionpost‐training

More Related Videos

Methods to Test Visual Attention Online
09:44

Methods to Test Visual Attention Online

Published on: February 19, 2015

12.7K
Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior
07:09

Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior

Published on: November 14, 2018

11.7K

Related Experiment Videos

Last Updated: Apr 9, 2026

Author Spotlight: Insights into the Analysis of Human Interaction with 3D Virtual Objects
06:36

Author Spotlight: Insights into the Analysis of Human Interaction with 3D Virtual Objects

Published on: October 18, 2024

1.5K
Methods to Test Visual Attention Online
09:44

Methods to Test Visual Attention Online

Published on: February 19, 2015

12.7K
Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior
07:09

Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior

Published on: November 14, 2018

11.7K

Area of Science:

  • Computer Vision
  • Robotics
  • Machine Learning

Background:

  • High-definition (HD) maps are crucial for autonomous driving perception and planning.
  • Current online HD map construction methods are computationally intensive, limiting real-time application on embedded systems.

Purpose of the Study:

  • To develop LiteMapNet, a computationally efficient transformer-based framework for accelerated online HD map construction.
  • To address the real-time deployment challenges of HD mapping on resource-constrained platforms.

Main Methods:

  • Proposed LiteMapNet, a lightweight transformer framework utilizing structured attention-head pruning.
  • Employed a two-stage pruning pipeline: Fisher information matrix for importance estimation and learnable masks for performance recovery.
  • Utilized lightweight calibration to enhance stability and generalization without inference overhead.

Main Results:

  • LiteMapNet significantly reduces computational cost and floating-point operations.
  • Achieved up to 1.5x inference speedup on NVIDIA H100 GPUs.
  • Maintained mapping accuracy with less than 1% drop compared to the unpruned baseline.

Conclusions:

  • LiteMapNet enables efficient online HD map construction for autonomous driving.
  • The framework facilitates deployment on resource-limited in-vehicle platforms.
  • The proposed pruning and calibration strategy balances efficiency and accuracy effectively.