Efficient convolutional neural networks and network compression methods for object detection: a survey

Object detection is one of the most basic and important research tasks in the field of computer vision. The general trend in object detection has been to design large and over-parameterized models, which can achieve excellent performance. However, this comes at the expense of low speed, heavy computation and large amount of memory overhead, also makes object detection models more difficult to be applied on mobiles and embedded devices which have limited hardware resources and need real-time feedback. So there has been rising interest in building portable and efficient networks for object detection in the recent literature. The main contributions of this review include the following aspects. As far as we know, there are few reviews on efficient object detection CNNs. We systematically summarize the methods, models and evaluation metrics of efficient CNNs for object detection in recent years. We summarize and introduce some commonly used datasets for object detection. Finally, we point out some possible research directions and inspire some useful suggestions for the future work of efficient convolutional neural network.