基于深度学习的目标检测研究综述

doi:10.3969/j.issn.0372-2112.2020.06.026

摘要/Abstract

摘要： 目标检测是计算机视觉领域内的热点课题，在机器人导航、智能视频监控及航天航空等领域都有广泛的应用.本文首先综述了目标检测的研究背景、意义及难点，接着对基于深度学习目标检测算法的两大类进行综述，即基于候选区域和基于回归算法.对于第一类算法，先介绍了基于区域的卷积神经网络（Region with Convolutional Neural Network，R-CNN）系列算法，然后从四个维度综述了研究者在R-CNN系列算法基础上所做的研究：对特征提取网络的改进研究、对感兴趣区域池化层的改进研究、对区域提取网络的改进研究、对非极大值抑制算法的改进研究.对第二类算法分为YOLO（You Only Look Once）系列、SSD（Single Shot multibox Detector）算法及其改进研究进行综述.最后根据当前目标检测算法在发展更高效合理的检测框架的趋势下，展望了目标检测算法未来在无监督和未知类别物体检测方向的研究热点.

关键词: 目标检测, 深度学习, 特征提取, 计算机视觉, 视频监控, 图像处理, 卷积神经网络

Abstract: Object detection is a hot topic in the field of computer vision, and has been widely used in robot navigation, intelligent video surveillance, aerospace, and other fields. The research background, significance and challenges of object detection were introduced. Then the object detection algorithms based on deep learning were reviewed according to two categories: candidate region-based and regression-based. For the candidate region-based algorithms, we first introduced the R-CNN (Region with Convolutional Neural Network) based series of algorithms, and then the R-CNN based methods were overviewed from four dimensions: the research of feature extraction networks, the region of interesting pooling researches, improved works based on region proposal networks, and some improved approaches of non maximum suppression algorithms. Next, the regression-based algorithms were surveyed in terms of YOLO (You Only Look Once) series and SSD (Single Shot multibox Detector) series. Finally, according to the current trend of object detection algorithms that are developing more efficient and reasonable detection frameworks, the future research focuses of unsupervised and unknown category object detection directions were prospected.

Key words: object detection, deep learning, feature extraction, computer vision, video surveillance, image processing, convolutional neural network

中图分类号:

TP391.4

罗会兰, 陈鸿坤. 基于深度学习的目标检测研究综述[J]. 电子学报, 2020, 48(6): 1230-1239.

LUO Hui-lan, CHEN Hong-kun . Survey of Object Detection Based on Deep Learning[J]. Acta Electronica Sinica, 2020, 48(6): 1230-1239.

参考文献

[1] LIU Li,OUYANG W,WANG Xiao-gang,et al.Deep Learning for Generic Object Detection:A Survey[OL].http://dblp.org/abs/1809.02165,2018.
[2] DALAL N,TRIGGS B.Histograms of oriented gradients for human detection[A].Proceedings of IEEE Conference on Computer Vision and Pattern Recognition[C].USA:IEEE,2005.886-893.
[3] LOWE D G.Distinctive image features from scale-invariant key points[J].International Journal of Computer Vision,2004,60(2):91-110.
[4] LIENHART R,MAYDT J.An extended set of haar-like features for rapid object detection[A].Proceedings of the International Conference on Image Processing[C].USA:IEEE,2002.900-903.
[5] SHAWE-TAYLOR J,CRISTIANINI N.An Introduction to Support Vector Machines and Other Kernel-based Learning Methods[M].England:Cambridge University Press,2000.
[6] FREUND Y,SCHAPIRE R E.Experiments with a new boosting algorithm[A].International Conference on Machine Learning[C].USA:IMLS,1996.148-156.
[7] LIAW A,WIENER M.Classification and regression by random-forest[J].R News,2002,2(3):18-22.
[8] 姜维,张重生,殷绪成.基于深度学习的场景文字检测综述[J].电子学报,2019,47(5):1152-1161. JIANG Wei,ZHANG Chong-sheng,YIN Xu-cheng.Deep learning based scene text detection:A survey[J].Acta Electronica Sinica,2019,47(5):1152-1161.(in Chinese)
[9] SHARMA K U,THAKUR N V.A review and an approach for object detection in images[J].International Journal of Computational Vision and Robotics,2017,7(1-2):196-237.
[10] 毕威,黄伟国,张永萍,等.基于图像显著轮廓的目标检测[J].电子学报,2019,45(8):1902-1910. BI Wei,HANG Wei-guo,ZHANG Yongping,et al.Object detection based on salient contour of image[J].Acta Electronica Sinica,2019,45(8):1902-1910.(in Chinese)
[11] CHAHAL KS,DEY K.A Survey of Modern Object Detection Literature Using Deep Learning[OL].http://dblp.org/abs/1808.07265,2018.
[12] OKSUZ Kemal,CAM Baris Can,KALKAN S,et al.Imbalance Problems in Object Detection:A Review[OL].http://dblp.org/abs/1909.00169,2019.
[13] ZHAO Z-Q,ZHENG P,XU S-T,et al.Object detection with deep learning:A review[J].IEEE Transactions on Neural Networks and Learning Systems,2019,30(11):3212-3232.
[14] REN S,HE K,GIRSHICK R,et al.Faster R-CNN:towards real-time object detection with region proposal networks[J].IEEE Trans Pattern Anal Mach Intell,2015,39(6):1137-49.
[15] GIRSHICK R,DONAHUE J,DARRELL T,et al.Rich feature hierarchies for accurate object detection and semantic segmentation[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2014.580-587.
[16] 周飞燕,金林鹏,董军.卷积神经网络研究综述[J].计算机学报,2017,40(6):1229-1251. ZHOU Fei-yan,JIN Lin-peng,DONG Jun.Review of convolutional neural network[J].Chinese Journal of Computers,2017,40(6):1229-1251.(in Chinese)
[17] KRIZHEVSKY A,SUTSKEVER I,HINTON G E.ImageNet classification with deep convolutional neural networks[J].Neural Information Processing Systems,2012,141(5):1097-1105.
[18] UIJLINGD J R,DE Sande K E,GEVERS T,et al.Selective search for object recognition[J].International Journal of Computer Vision,2013,104(2):154-71.
[19] EVERINGHAM M,VAN GOOL L,WILLIAMS C K,et al.The pascal visual object classes (VOC) challenge[J].International Journal of Computer Vision,2010,88(2):303-38.
[20] HE K,ZHANG X,REN S,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[A].Proceedings of the European Conference on Computer Vision[C].Switzerland:IEEE,2014.346-361.
[21] GIRSHICK R.Fast R-CNN[A].Proceedings of IEEE International Conference on Computer Vision[C].USA:IEEE,2015.1440-1448.
[22] LONG J,SHELHAMER E,DARRELL T.Fully convolutional networks for semantic segmentation[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2015.3431-3440.
[23] NEUBECK A,VAN GOOL L.Efficient non-maximum suppression[A].Proceedings of the International Conference on Pattern Recognition[C],USA:IEEE,2006.850-855.
[24] KONG T,YAO A,CHEN Y,et al.Hypernet:towards accurate region proposal generation and joint object detection[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2016.845-853.
[25] 黄继鹏,史颖欢,高阳.面向小目标的多尺度Faster-RCNN检测算法[J].计算机研究与发展,2019,56(2):319-327. HUANG Ji-peng,SHI Ying-huan,GAO Yang.Multi-scale faster R-CNN algorithm for small object detection[J].Journal of Computer Research and Development,2019,56(2):319-327.(in Chinese)
[26] LIN T Y,DOLLAR P,GIRSHICK R,et al.Feature pyramid networks for object detection[A].Proceedings of IEEE Conference on Computer Vision and Pattern Recognition[C].USA:IEEE,2017.936-944.
[27] SINGH B,DAVIS L S.An analysis of scale invariance in object detection-snip[A].Proceedings of IEEE Conference on Computer Vision and Pattern Recognition[C].USA:IEEE,2018.3578-3587.
[28] DAI J,LI Y,HE K,et al.R-FCN:object detection via region-based fully convolutional networks[A].Neural Information Processing Systems[C].Spain:NIPS Foundation,2016.379-387.
[29] ZHU Y,ZHAO C,WANG J,et al.Couplenet:coupling global structure with local parts for object detection[A].the IEEE International Conference on Computer Vision[C].Italy:IEEE,2017.4146-4154.
[30] DAI J,QI H,XIONG Y,et al.Deformable convolutional networks[A].International Conference on Computer Vision,Venice[C].Italy:IEEE,2017.764-773.
[31] ZHU X,HU H,LIN S,et al.Deformable Convnets-v2:More Deformable,Better Results[OL].http://dblp.org/abs/1811.11168,2018.
[32] HE K,GKIOXARI G,DOLLAR P,et al.Mask r-cnn[A].International Conference on Computer Vision[C].Italy;IEEE,2017.2980-2988.
[33] JIANG B,LUO R,MAO J,et al.Acquisition of localization condence for accurate object detection[A].European Conference on Computer Vision[C].Germany:IEEE,2018.816-832.
[34] CAI Z,VASCONCELOS N.Cascade R-CNN:delving into high quality object detection[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2018.
[35] CHEN Z,HUANG S,TAO D.Context refinement for object detection[A].Proceedings of the European Conference on Computer Vision[C].Germany:IEEE,2018.6154-6162.
[36] WANG J,CHEN K,YANG S,et al.Region proposal by guided anchoring[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2019.2965-2974.
[37] BODLA N,SINGH B,CHELLAPPA R,et al.Soft-NMS:improving object detection with one line of code[A].International Conference on Computer Vision[C].Italy:IEEE,2017.5562-5570.
[38] HE Yi-hui,ZHANG Xiang-yu,MARIOS S,et al.Softer-NMS:Rethinking Bounding Box Regression for Accurate Object Detection[OL].http://dblp.org/abs/1809.08545,2018.
[39] HU H,GU J,ZHANG Z,et al.Relation networks for object detection[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2018.3588-3597.
[40] VASWANI A,SHAZEER N,PARMAR N,et al.Attention is all you need[A].Neural Information Processing Systems[C].USA:NIPS Foundation,2017.5998-6008.
[41] REZATOFIGHI S H,TSOI N,GWAK J,et al.Generalized Intersection over Union:A Metric and A Loss for Bounding Box Regression[OL].http://dblp.org/abs/1902.09630,2019.
[42] REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once:unified,real-time object detection[A].IEEE Conference on Computer Vision and Pattern Recognition[C].USA:IEEE,2016.779-778.
[43] LIU W,ANGUELOV D,ERHAN D,et al.SSD:single shot multibox detector[A].European Conference on Computer Vision,Amsterdam[C].Netherlands:IEEE,2016.21-37.
[44] REDMON J,FARHADI A.YOLO9000:Better,faster,stronger[A].Computer Vision and Pattern Recognition,Hawaii[C].USA:IEEE,2017.6517-6525.
[45] RUSSAKOVSKY O,DENG J,SU H,et al.Imagenet large scale visual recognition challenge[J].International journal of computer vision,2015,115(3):211-52.
[46] LIN T,MAIRE M,BELONGIE S J,et al.Microsoft coco:common objects in context[A].European Conference on Computer Vision[C].Switzerland:IEEE,2014.740-755.
[47] REDMON J,FARHADI A.YOLOV3:An incremental improvement[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2018.3523-3541.
[48] HE K,ZHANG X,REN S,et al.Deep residual learning for image recognition[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2016.770-778.
[49] SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[A].International Conference on Learning Representations[C].USA:IEEE,2015.714-723.
[50] JEONG J,PARK H,KWAK N.Enhancement of SSD by concatenating feature maps for object detection[A].British Machine Vision Conference[C].UK:BMVA,2017.
[51] FU Cheng-yang,LIU Wei,RANGA A,et al.DSSD:Deconvolutional Single Shot Detector[OL].http://arxiv.org/abs/1701.06659,2017.
[52] LIN T-Y,GOYAL P,GIRSHICK R,et al.Focal loss for dense object detection[A].International Conference on Computer Vision[C].USA:IEEE,2017.2999-3007.
[53] LI B,LIU Y,WANG X.Gradient harmonized single-stage detector[A].the AAAI Conference on Artificial Intelligence[C].USA:IEEE,2019.8577-8584.
[54] LIU S,HUANG D,WANG Y.Receptive field block net for accurate and fast object detection[A].European Conference on computer Vision[C].Germany:IEEE,2018.404-419.
[55] SZEGEDY C,VANHOUCKE V,IOFFE S,et al.Rethinking the Inception Architecture for Computer Vision[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2016.2818-2826.
[56] ZHANG S,WEN L,BIAN X,et al.Single shot refinement neural network for object detection[A].IEEE Conference on Computer Vision and Pattern Recognition[C].USA:IEEE,2018.4203-4212.
[57] SUN F,KONG T,HUANG W,et al.Feature pyramid reconfiguration with consistent loss for object detection[J].IEEE Transactions on Image Processing,2019,28(10):5041-5051.
[58] GHIASI G,LIN T-Y,LE Q V.NAS-FPN:learning scalable feature pyramid architecture for object detection[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2019.7036-7045.
[59] ZHAO Q,SHENG T,WANG Y,et al.M2det:A single-shot object detector based on multi-level feature pyramid network[A].The AAAI Conference on Artificial Intelligence[C].USA:IEEE,2019.9259-9266.
[60] 裴伟,许晏铭,朱永英,等.改进的SSD航拍目标检测方法[J].软件学报,2019,30(3):738-758. FEI Wei,XU Yan-ming,ZHU Yong-ying,et al.The target detection method of aerial photography images with Improved SSD[J].Journal of Software,2019,30(3):738-758.(in Chinese)
[61] YI J,WU P,METAXAS D N.ASSD:Attentive single shot multibox detector[J].Computer Vision and Image Understanding,2019,189(1):102827-102836.
[62] LAW H,DENG J.Cornernet:detecting objects as paired keypoints[A].European Conference on Computer Vision[C].Germany:IEEE,2018.765-781.
[63] NEWELL A,YANG K,DENG J.Stacked hourglass networks for human pose estimation[A].European Conference on Computer Vision[C].Netherlands:IEEE,2016.483-499.
[64] ZHOU Xing-yi,ZHUO Jia-cheng,KRAHENBUHL P.Bottom-up object detection by grouping extreme and center points[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2019.850-859.
[65] DUAN Kai-wen,BAI Song,XIE Ling-xi,et al.Centernet:Keypoint Triplets for Object Detection[OL].http://dblp.org/abs/1904.08189,2019.
[66] TIAN Zhi,SHEN Chun-hua,CHEN Hao,et al.FCOS:Fully Convolutional One-Stage Object Detection[OL].http://dblp.org/abs/1904.01355,2019.
[67] SHETTY S.Application of convolutional neural network for image classification on pascal VOC challenge 2012 dataset[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2016.
[68] KUZNETSOVA A,ROM H,ALLDRIN N,et al.The Open Images Dataset-v4:Unified Image Classification,Object Detection,and Visual Relationship Detection at Scale[OL].http://dblp.org/abs/1811.00982,2018.
[69] GUPTA A,DOLLAR P,GIRSHICK R.LVIS:a dataset for large vocabulary instance segmentation[A].Computer Vision and Pattern Recognition[C].USA:IEEE,2019.5356-5364.