Multi-object Detection Method based on YOLO and ResNet Hybrid Networks

As a sub-discipline related to artificial intelligence, machine vision has become a hotspot in recent years. Object detection, as part of machine vision, has been widely used in various fields. However, the existing object detection methods based on deep learning theory cannot accurately recognize multiple objects. Thus, in this paper we propose a new hybrid network based on YOLO and ResNet (Yolo-resnet) for multi-object detection. Our method integrates ResNet into the feature extraction of the YOLOv3 framework and the detection results demonstrate that our hybrid network is efficient for detecting multi-objects from an image of complex natural scenes.