AN OVERVIEW OF IMAGE RECOGNITION AND REAL-TIME OBJECT DETECTION

Abstract

Nowadays neural networks, deep learning and computer vision give us the best solutions for many problems in Artificial intelligence such as image recognition, speech recognition, natural language processing, price and load forecasting, healthcare, marketing, recommendation systems, etc. This paper presents an overview of research in the field of image recognition done in recent years. First an outline of object recognition and deep learning methodologies are given. Then are explained convolutional neural networks as architecture for deep learning.  In addition, image annotation techniques and tools are evaluated, in order to choose those with the best performance and accuracy. At the end are discussed the frameworks for real-time object detection in prospect of our research.