Research

Share    

Extreme learning machines based image classification and scene characterization for autonomous driving

Objective

To develop an efficient Extreme Learning Machine (ELM) based solution for object classification and detection during autonomous driving.  


Description

The project team has developed two state-of-the-art object detectors for performance analysis: namely Faster Region based Convolutional Neural Network (R-CNN) and Hierarchical- Extreme Learning Machine (H-ELM) based object detector. Attributed to its superiority in feature learning and object localisation, the detection performance delivered by Faster R-CNN is more promising compared to H-ELM.Despite the satisfactory result produced by R-CNN, the implementation of the algorithm for practical online application is still limited by its tedious computational effort. The team is currently looking into the possibility of combining both algorithms in a hybrid framework in order to obtain a trade-off between detection performance and computational efficiency. 

 

                                   Object classification using Convolutional Neural Network (CNN)                                                                   and Hierarchical-Extreme Learning Machine ( H-ELM) in a hybrid framework. 


Result

      
                                                                     Real-time object detection



Demonstration Video






Principal Investigators


Prof. Huang Guangbin (NTU)

Telephone: 6790 4489
Office: S2.1-B2-06
Email: egbhuang@ntu.edu.sg
Mr. Paul Tan (STE)

Telephone: 6660 1052
Office: S1-B4a-03
Email: paultan@stengg.com