Review of Object Detection Techniques

As the front-end technology of artificial intelligence, computer vision has been widely studied in recent years, and the introduction of deep learning methods has accelerated this process. This paper shows the progress made in object detection in the last 5 years, followed by the mainstream model topology including Convolutional Neural Network and Transformer. We further compared the accuracy and model complexity of different backbones, analyzed the differences and the inner link between Convolutional Neural Network and Transformer, at the end of the thesis, the prospect of future development is presented.