Search research articles

Related Concept Videos

Improving Translational Accuracy

Improving Translational Accuracy

You might also read

Related Articles

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

Sort by

Same author

Quaternion-Based Attitude Estimation of an Aircraft Model Using Computer Vision.

Sensors (Basel, Switzerland)·2024

Same author

Enhancing Aircraft Safety through Advanced Engine Health Monitoring with Long Short-Term Memory.

Sensors (Basel, Switzerland)·2024

Same author

Semantic Terrain Segmentation in the Navigation Vision of Planetary Rovers-A Systematic Literature Review.

Sensors (Basel, Switzerland)·2022

Same author

Drone Model Classification Using Convolutional Neural Network Trained on Synthetic Data.

Journal of imaging·2022

Same author

Sky and Ground Segmentation in the Navigation Visions of the Planetary Rovers.

Sensors (Basel, Switzerland)·2021

Same journal

Human-AI Interaction in Interventional Radiology: A Narrative Review of Current Applications, Challenges, and Future Directions.

Journal of imaging·2026

Same journal

Coronary Artery Anomalies and Anatomical Variants: Cross-Sectional Diagnostic Imaging and Clinical Background.

Journal of imaging·2026

Same journal

YoLeTooth: A Unified Framework for Joint Tooth Segmentation and Periapical Lesion Detection in Panoramic Radiographs.

Journal of imaging·2026

Same journal

Radiomics-Guided Multi-Sequence Learning for Pathological Complete Response Prediction from Breast MRI with Missing Auxiliary Sequences.

Journal of imaging·2026

Same journal

Cutaneous Thermography in Arthropathies: Quantitative Imaging, Machine Learning, and Clinical Translation.

Journal of imaging·2026

Same journal

Two-Stage Dynamic Synergistic Segmentation Method for Myocardial Pathology.

Journal of imaging·2026

See all related articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Jul 10, 2025

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

Constraints on Optimising Encoder-Only Transformers for Modelling Sign Language with Human Pose Estimation Keypoint

Luke T Woods^1,2, Zeeshan A Rana³

¹Digital Aviation Research and Technology Centre (DARTeC), Cranfield University, Cranfield, Bedfordshire MK43 0AL, UK.

Journal of Imaging

|November 24, 2023

Summary

This summary is machine-generated.

Regularisation techniques offer limited benefits for sign language deep learning models, with only L2 regularization showing impact. Dataset size significantly bounds model performance, despite efforts to optimize hyperparameters.

Keywords:

classification computer vision data augmentation deep learning human pose estimation machine learning regularisation sign language recognition supervised learning

More Related Videos

Author Spotlight: Automated Deep Brain Stimulation for Parkinson's Disease - Exploring the Possibilities and Challenges of Home Monitoring

Author Spotlight: Automated Deep Brain Stimulation for Parkinson's Disease - Exploring the Possibilities and Challenges of Home Monitoring

Published on: July 14, 2023

Author Spotlight: Deciphering Electrical Networks Behind Complex Brain Activities and Disorders

Author Spotlight: Deciphering Electrical Networks Behind Complex Brain Activities and Disorders

Published on: November 1, 2024

Related Experiment Videos

Last Updated: Jul 10, 2025

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

Author Spotlight: Automated Deep Brain Stimulation for Parkinson's Disease - Exploring the Possibilities and Challenges of Home Monitoring

Author Spotlight: Automated Deep Brain Stimulation for Parkinson's Disease - Exploring the Possibilities and Challenges of Home Monitoring

Published on: July 14, 2023

Author Spotlight: Deciphering Electrical Networks Behind Complex Brain Activities and Disorders

Author Spotlight: Deciphering Electrical Networks Behind Complex Brain Activities and Disorders

Published on: November 1, 2024

Area of Science:

Computer Science
Artificial Intelligence
Machine Learning

Background:

Supervised deep learning models require regularization to prevent overfitting, a process complicated by hyperparameter tuning.
Understanding the influence of individual hyperparameters and regularization techniques is crucial for optimizing model performance in research.

Purpose of the Study:

To conduct a comprehensive, large-scale ablation study on an encoder-only transformer for sign language modeling.
To assess the impact of various regularization and data augmentation techniques on sign classification accuracy.
To identify performance constraints and optimize the sign language modeling task.

Main Methods:

Utilized the improved Word-level American Sign Language dataset (WLASL-alt) and human pose estimation keypoint data.
Performed a large-scale ablation study on an encoder-only transformer architecture.
Measured the effect of diverse model parameter regularization and data augmentation techniques on classification accuracy.

Main Results:

Except for L2 parameter regularization, tested techniques showed no significant positive impact on performance, contradicting some prior studies.
Model architecture performance was found to be constrained by the limited dataset size.
Achieved a new benchmark top-1 classification accuracy of 84% on 100 signs using the base model configuration.

Conclusions:

The effectiveness of common regularization and data augmentation techniques is limited for this specific sign language modeling task.
Dataset size is a critical bottleneck for achieving higher performance in sign language classification with current architectures.
The study provides valuable insights into hyperparameter optimization and sets a new performance benchmark for the WLASL-alt dataset.