Pedestrian Detection Using Deep Residual Network with Transfer Learning and Faster RCNN

With the rise of Artificial Intelligence over the past decade, many useful applications have been developed and in use. Computer vision is one of the core areas of Machine Learning. Computer vision could solve a range of tasks in many areas like medical computer vision, manufacturing machine vision, in autonomous vehicles and so on. Pedestrian detection is a challenging problem in computer vision. It is also an important task in video monitoring system. The objective of this thesis is to build a true pedestrian detector. The robust model is based on deep residual neural network using Transfer Learning. Trained model can be used to detect pedestrians on traffic videos in real time.