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: Jul 5, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.2K

Deeply Supervised Block-Wise Neural Architecture Search.

An Yang, Ying Liu, Chunguang Li

    IEEE Transactions on Neural Networks and Learning Systems
    |January 17, 2024
    PubMed
    Summary

    We introduce Deeply Supervised Block-wise Neural Architecture Search (DBNAS), a resource-friendly method for designing neural networks. DBNAS efficiently searches for promising architectures with significantly reduced computational cost and memory footprint.

    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

    Resolving Thermal Accumulation and Rigid-Soft Interface Mismatch in Stretchable Electronics with Cubic Boron Nitride Composite Islands.

    ACS applied materials & interfaces·2026
    Same author

    Cellulose-Based Superabsorbent Hydrogel with Recyclable Moisture Regulation for Agriculture in Arid Regions.

    Journal of agricultural and food chemistry·2026
    Same author

    Efficacy and safety of ozone autohemotherapy for zoster-associated pain: a meta-analysis and trial sequential analysis.

    Frontiers in neurology·2026
    Same author

    Class prototype rectification and multi-scale feature measurement for few-shot classification of bearing surface defects.

    Scientific reports·2026
    Same author

    Leaf- and root-associated bacterial communities differ in their resistance and resilience to N disturbance in a temperate steppe.

    Applied and environmental microbiology·2026
    Same author

    Serum cystatin C levels are independently correlated with cognitive impairment in individuals with cerebral small vessel disease.

    Frontiers in neuroscience·2026

    Area of Science:

    • Artificial Intelligence
    • Machine Learning
    • Computer Vision

    Background:

    • Neural Architecture Search (NAS) automates neural network design.
    • Block-wise NAS addresses weight-sharing limitations but incurs high computational costs.
    • Existing methods use supervised distillation or contrastive learning, demanding extensive resources.

    Purpose of the Study:

    • To propose a resource-friendly block-wise NAS method.
    • To alleviate the computational burden of current block-wise NAS techniques.
    • To enable efficient and effective neural network architecture optimization.

    Main Methods:

    • Developed Deeply Supervised Block-wise NAS (DBNAS).
    • Integrated lightweight deeply-supervised modules after each network block.

    More Related 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

    556
    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
    04:48

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

    Published on: July 5, 2024

    405

    Related Experiment Videos

    Last Updated: Jul 5, 2025

    Deep Neural Networks for Image-Based Dietary Assessment
    13:19

    Deep Neural Networks for Image-Based Dietary Assessment

    Published on: March 13, 2021

    9.2K
    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

    556
    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
    04:48

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

    Published on: July 5, 2024

    405
  • Employed a simple supervised learning scheme using ground-truth labels for progressive block optimization.

    • Main Results:

    • DBNAS achieved efficient architecture search on ImageNet in under 1 GPU day.
    • Required less GPU memory compared to existing block-wise NAS methods.
    • The best DBNAS model reached 75.6% Top-1 accuracy on ImageNet, competitive with state-of-the-art.

    Conclusions:

    • DBNAS offers a computationally efficient and effective approach to block-wise NAS.
    • The method demonstrates strong performance on ImageNet and good transferability to other datasets and tasks.
    • DBNAS successfully balances search efficiency with model performance.