Addressing Computer Vision Challenges Using an Active Learning Framework

Abstract

Computer vision has introduced new successful opportunities in everyday life. Recently, there has been a lot of progress, particularly in face recognition and object detection systems. These systems require a large amount of data to be trained with, in order to perform satisfyingly. Active learning addresses this challenge by leveraging a small amount of manually labelled data. This paper builds on state-of-the-art face recognition and object detection models, by implementing optimization techniques that enhance the recognition accuracy. Further training is being introduced by making use of a robust active learning framework that results in creating extended data sets. Finally, the paper proposes an integrated system, which involves efficient techniques of associating face and object identification information, in order to extract (in real-time) as much knowledge as possible from a video streaming.