改进多尺度卷积结构与高斯核的E-CenterNet算法

doi:10.3778/j.issn.1002-8331.2205-0430

摘要/Abstract

摘要： 针对ResNet和DLA（deep layer aggregation）为主干网络的CenterNet算法特征提取能力不充分、热力图与目标真实边界框吻合度不高、关键点损失函数未充分考虑预测值对难易样本训练比重的影响等不足，提出一种改进多尺度卷积结构与高斯核的E-CenterNet算法。引入轻量型EfficientNetV2-S作为主干网络，并结合基于金字塔分割注意力网络的多尺度卷积结构对其进行改进，提高特征提取能力；对高斯核进行改进，使CenterNet产生的热力图由固定的圆形改进为随边界框宽高变化的椭圆形，增强算法对边界框宽高差异大的目标的检测能力；提出一种基于关键点预测值的关键点损失函数，提高算法对难样本的训练比例。在Pascal VOC数据集上的实验结果表明：E-CenterNet算法的mAP达到83.3%，比原始算法提升了2.6个百分点，检测性能优于CenterNet算法。

关键词: CenterNet, 目标检测, 多尺度卷积, 高斯核, 关键点损失函数

Abstract: As the detection algorithm, the CenterNet based on ResNet and DLA（deep layer aggregation） as the backbone doesn’t have sufficient capabilities for feature extractions, neither its heatmap accords with the target real bounding box nor the keypoint loss function fully consider the impact of the predicted value on the training proportion of difficult and easy samples. In order to solve these deficiencies, E-CenterNet with more advanced multi-scale convolution structure and Gaussian kernel is proposed. Firstly, lightweight EfficientNetV2-S is introduced as the backbone, and the multi-scale convolution structure based on the pyramid split attention network is used to enhance the feature extraction ability. Secondly, by improving the Gaussian kernel, the heatmap generated by CenterNet is enhanced from a fixed circle to an ellipse varying with the width and height of the bounding box, which enhances the detection ability of algorithm for objects with large differences in the width and height of the bounding box. Finally, the Keypoint loss function based on the predicted value is proposed in order to augment the training ratio of the algorithm E-CenterNet for difficult samples. The experimental results on the Pascal VOC indicate that the mAP of the E-CenterNet accounts for 83.3%, which is 2.6 percentage points higher than that of the CenterNet, and the detection performance of E-CenterNet is better than that of the CenterNet.

Key words: CenterNet, object detection, multi-scale convolution, Gaussian kernel, keypoint loss function

胡松松, 吴亮红, 张红强, 陈亮, 周博文, 张侣. 改进多尺度卷积结构与高斯核的E-CenterNet算法[J]. 计算机工程与应用, 2023, 59(6): 70-80.

HU Songsong, WU Lianghong, ZHANG Hongqiang, CHEN Liang, ZHOU Bowen, ZHANG Lyu. E-CenterNet Algorithm with Improved Multi-Scale Convolution Structure and Gaussian Kernel[J]. Computer Engineering and Applications, 2023, 59(6): 70-80.

参考文献

[1] VIOLA P，JONES M J.Robust real-time face detection[J].International Journal of Computer Vision，2004，57（2）：137-154.
[2] DALAL N，TRIGGS B.Histograms of oriented gradients for human detection[C]//2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition（CVPR’05），2005：886-893.
[3] FELZENSZWALB P，MCALLESTER D，RAMANAN D.A discriminatively trained，multiscale，deformable part model[C]//2008 IEEE Conference on Computer Vision and Pattern Recognition，2008：1-8.
[4] LOWE D G.Distinctive image featuresfrom scale-invariant keypoints[J].International Journal of Computer Vision，2004，60（2）：91-110.
[5] BAY H，TUYTELAARS T，GOOL L V.SURF：speeded up robustfeatures[J].Computer Vision & Image Understanding，2006，110（3）：404-417.
[6] REDMON J，DIVVALA S，GIRSHICK R，et al.You only look once：unified，real-time object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2016：779-788.
[7] REDMON J，FARHADI A.YOLO9000：better，faster，stronger[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2017：7263-7271.
[8] WANG K，LIU M Z.YOLOv3-MT：a YOLOv3 using multi-target tracking for vehicle visual detection[J].Applied Intelligence，2022，52（2）：2070-2091.
[9] YU J M，ZHANG W.Face mask wearing detection algorithm based on improved YOLO-v4[J].Sensors，2021，21（9）：3263-3283.
[10] 马玉莹，黄成章，黄静颖，等.基于SSD框架的红外弱小目标检测技术研究[J].激光与红外，2021，51（10）：1342-1347.
MA Y Y，HUANG C Z，HUANG J Y，et al.Research of infrared dim and small targets detection technology based on SSD framework[J].Laser and Infrared，2021，51（10）：1342-1347.
[11] ARAKI R，HIRAKAWA T，YAMASHITA T，et al.MT-DSSD：multi-task deconvolutional single shot detector for object detection，segmentation，and grasping detection[J].Advanced Robotics，2022，36（8）：373-387.
[12] BROMAN G I，ROBèRT K H.A framework for strategic sustainable development[J].Journal of Cleaner Production，2017，140（1）：17-31.
[13] LAW H，DENG J.CornerNet：detecting objects as paired keypoints[J].International Journal of Computer Vision，2020，128（2）：734-750.
[14] ZHOU X Y，ZHUO J C，KRAHENBUHL P.Bottom-up object detection by grouping extreme and center points[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition，2019：850-859.
[15] XIE H S，WU Z S.A robust fabric defect detection method based on improved RefineDet[J].Sensors，2020，20（15）：4260.
[16] DUAN K W，BAI S，XIE L X，et al.Centernet：keypoint triplets for object detection[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision，2019：6569-6578.
[17] 宫大汉，于龙龙，陈辉，等.面向车规级芯片的对象检测模型优化方法[J].智能系统学报，2021，16（5）：900-907.
GONG D H，YU L L，CHEN H，et al.Object detection model optimization method for car-level chips[J].CAAI Transactions on Intelligent Systems，2021，16（5）：900-907.
[18] 魏玮，杨茹，朱叶.改进CenterNet的遥感图像目标检测[J].计算机工程与应用，2021，57（6）：191-199.
WEI W，YANG R，ZHU Y.Target detection of improved centernet to remote sensing images[J].Computer Engineering and Applications，2021，57（6）：191-199.
[19] GIRSHICK R，DONAHUE J，DARRELL T，et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2014：580-587.
[20] HE K M，ZHANG X Y，REN S Q，et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2015，37（9）：1904-1916.
[21] YIN G X，YU M，WANG M，et al.Research on highway vehicle detection based on faster R-CNN and domain adaptation[J].Applied Intelligence，2022，52（4）：3483-3498.
[22] REN S Q，HE K M，GIRSHICK R，et al.Faster R-CNN：towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2017，39（6）：1137-1149.
[23] WANG J L，LUO J X，LIU B，et al.Automated diabetic retinopathy grading and lesion detection based on the modified R-FCN object‐detection algorithm[J].IET Computer Vision，2020，14（1）：1-8.
[24] HE K M，ZHANG X Y，REN S Q，et al.Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2016：770-778.
[25] NEWELl A，YANG K Y，DENG J.Stacked hourglass networks for human pose estimation[C]//European Conference on Computer Vision.Cham：Springer，2016：483-499.
[26] 赵锐，赵国伟，张娟，等.改进CenterNet的高压输电线路巡检故障实时检测方法[J].计算机工程与应用，2021，57（17）：246-252.
ZHAO R，ZHAO G W，ZHANG J，et al.Real-time faultdetection method for high voltage transmission line based on CenterNet improved algorithm[J].Computer Engineering and Applications，2021，57（17）：246-252.
[27] TAN M M，LE Q.Efficientnetv2：Smaller models and faster training[C]//International Conference on Machine Learning，2021：10096-10106.
[28] ZHANG H，ZU K K，LU J，et al.EPSANet：an efficient pyramid squeeze attention block on convolutional neural network[J].arXiv：2105.14447，2021.
[29] EVERINGHAM M，WINN J.The PASCAL visual object classes challenge 2007 （VOC2007） development kit[J].International Journal of Computer Vision，2010，88（2）：303-338.
[30] EVERINGHAM M，ESLAMI S M，VAN G L，et al.The pascal visual object classes challenge：a retrospective[J].International Journal of Computer Vision，2015，111（1）：98-136.
[31] WANG Q L，WU B G，ZHU P F，et al.ECA-Net：efficient channel attention for deep convolutional neural networks[J].arXiv：1910.03151，2019.
[32] LEE H J，KIM H E，NAM H.SRM：a style-based recalibration module for convolutional neural networks[J].arXiv：1903.10829，2019.