Object Interaction Detection Based on Convolutional Neural Network

Abstract

Vision gives people a strong recognition ability, and people can easily see various objects. With the popularity of smart phones and the development of the Internet, pictures and videos have become an important way for people to record their lives and share information. They are created so fast that no one can search them all, but the wealth of information they contain is very important to people. Therefore, this article is based on the convolutional neural network to study the object interaction detection. First of all, this article discusses the basic concepts of convolutional neural networks, and conducts corresponding research on the convolution algorithm, and then studies the method of object interaction detection. After that, the performance of traditional neural network and convolutional neural network in object interaction detection is tested and researched. The test results show that the convolutional neural network is faster than the traditional neural network algorithm in the four types of performance of image generation speed, object feature extraction speed, object classification speed, and object detection accuracy.