Real-time driver action recognition system
Date of Issue2016
School of Electrical and Electronic Engineering
Through recent years, the interest of research in the area of human action recognition in computer vision has increased rapidly. The main objective of this FYP is to develop a system for recognizing different gestures. The most important tool used is Extreme Machine Learning Algorithm, because of its fast speed for analyzing. Two tests were conducted after the training process. The test with samples collected from real-time webcam achieves an accuracy of 50%, while the test with samples from original training database is much higher. Future work could be focused on improving the accuracy and apply the system to a wider field such as whole-body action recognition.
Final Year Project (FYP)
Nanyang Technological University