Related Experiment Video
Updated: Oct 7, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
674
iELMNet: Integrating Novel Improved Extreme Learning Machine and Convolutional Neural Network Model for Traffic Sign
Aisha Batool1, Muhammad Wasif Nisar1, Jamal Hussain Shah1
1Department of Computer Science, COMSATS University Islamabad, Wah Campus, Islamabad, Pakistan.
Big Data
|January 7, 2022
Summary
This study introduces an improved extreme learning machine network for real-time traffic sign detection (TSD). The novel model enhances accuracy and efficiency in automated driving systems by integrating advanced deep learning techniques.
Area of Science:
- Computer Vision
- Machine Learning
- Artificial Intelligence
Background:
- Real-time traffic sign detection (TSD) is crucial for automated driving systems.
- High intraclass variation due to diverse sign appearances and spatial representations poses challenges.
Purpose of the Study:
- To propose an efficient and accurate real-time traffic sign detection model.
- To address the challenges of intraclass variation and varying sign sizes.
Main Methods:
- Developed an improved extreme learning machine network (ELM) model.
- Utilized a custom DenseNet-based convolutional neural network (CNN) architecture.
- Incorporated an accurate anchor prediction model (A2PM), scale transformation (ST), and ELM module.
Main Results:
- Achieved high average precisions: 93.31% on Tsinghua-Tencent 100K, 95.22% on unreal/real environments, and 99.45% on German traffic sign detection benchmark.
- Demonstrated superior performance compared to existing state-of-the-art sign detection techniques.
Conclusions:
- The proposed ELM network model offers enhanced efficiency and accuracy for real-time TSD.
- The integration of DenseNet-CNN, A2PM, ST, and ELM effectively handles intraclass variations and scale differences.
