Abstract
Faster Region-based Convolutional Network (Faster R-CNN) is a state-of-the-art object detection method. However, the object detection effect of Faster R-CNN is not good based on the Region Proposal Network (RPN). Inspired by RPN of Faster R-CNN, we propose a novel proposal generation method called Enhanced Region Proposal Network (ERPN). Four improvements are presented in ERPN. Firstly, our proposed deconvolutional feature pyramid network (DFPN) is introduced to improve the quality of region proposals. Secondly, novel anchor boxes are designed with interspersed scales and adaptive aspect ratios. Thereafter, the capability of object localization is increased. Thirdly, a particle swarm optimization (PSO) based support vector machine (SVM), termed PSO-SVM, is developed to distinguish the positive and negative anchor boxes. Fourthly, the classification part of multi-task loss function in RPN is improved. Consequently, the effect of classification loss is strengthened. In this study, our proposed ERPN is compared with five object detection methods on both PASCAL VOC and COCO data sets. For the VGG-16 model, our ERPN obtains 78.6% mAP on VOC 2007 data set, 74.4% mAP on VOC 2012 data set and 31.7% on COCO data set. The performance of ERPN is the best among the comparison object detection methods. Furthermore, the detection speed of ERPN is 5.8 fps. Additionally, ERPN obtains good effect on small object detection.
Highlights
The object detection [1,2,3,4,5] problems are one of the key tasks in the computer vision field
Inspired by RPN of Faster Region-based Convolutional Neural Network (R-CNN), we propose a novel proposal generation method called Enhanced Region Proposal Network (ERPN)
Novel anchor boxes are designed with interspersed scales and adaptive aspect ratios
Summary
The object detection [1,2,3,4,5] problems are one of the key tasks in the computer vision field. Region proposals are applied by most of the top-performing object detection methods to search for objects. The reason is that Fast R-CNN combines the region proposal classification and bounding box regression tasks into one single stage. Because SS method is applied to generate region proposals in Fast R-CNN, thereafter the detection speed of Fast R-CNN is affected. Two processes of Faster R-CNN are presented as follows: First, SS method is replaced by RPN which is a kind of fully convolutional network [17], [18] (FCN) and can be trained end-to-end to generate detection proposals. Inspired by RPN of Faster R-CNN, we propose a novel proposal generation method called Enhanced Region Proposal Network (ERPN). The performance of small object detection is promoted by applying the rich top-level features. The performance of our ERPN based Faster R-CNN method is outstanding
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
