基于无锚框分割网络改进的实例分割方法

doi:10.19678/j.issn.1000-3428.0062846

摘要/Abstract

摘要： 在无人驾驶应用场景中，现有无锚框实例分割方法存在大目标特征覆盖小目标特征、缺少两阶段检测器中的感兴趣区域对齐操作、忽略类别分支对掩膜分支提供的位置和空间信息等问题，导致特征提取不充分且无法准确获取目标区域。提出一种改进的无锚框实例分割方法。结合可变形卷积，设计编码-解码特征提取网络提取高分辨率特征，以增强对小目标特征的提取能力，并采用空洞卷积和合并连接的方式，在不增加计算量的前提下有效融合多种分辨率的特征。在此基础上，将注意力机制引入到类别分支中，同时设计结合空间信息和通道信息的信息增强模块，以提高目标检测能力。实验结果表明，该方法在COCO 2017和Cityscapes数据集上平均精度和平均交并比分别为41.1%和83.3%，相比Mask R-CNN、SOLO、Yolact等方法，能够有效改进实例分割效果并具有较优的鲁棒性。

关键词: 无锚框实例分割, 深度学习, 编码-解码结构, 注意力机制, 空洞卷积

Abstract: In autonomous driving application scenarios, the existing anchor-free instance segmentation methods have problems such as large target features covering small target features, lack of a Region Of Interest (ROI)-Align operation in the two-stage detector, ignoring the position and spatial information provided by the regression branch to the mask branch, resulting in insufficient feature extraction and unable to accurately obtain the target region.An improved method for instance, anchor-free segmentation is proposed.Combined with deformable convolution, a encoder-decoder feature extraction network is designed to extract high-resolution features and enhance the extraction ability of small target features.The dilated convolution and merging connection method is adopted to effectively fuse the features of multiple resolutions without increasing the computation amount.On this basis, the attention mechanism is introduced into the regression branch, and an information enhancement module combining spatial and channel information is designed to improve the ability of target detection.The experimental results show that the Average Precision(AP) and mean Intersection over Union(mIoU) of the proposed method on the COCO 2017 and Cityscapes datasets are 41.1% and 83.3%, respectively.Compared with Mask R-CNN, SOLO, Yolact, and other methods, the proposed method can effectively improve the effect of instance segmentation and has better robustness.

Key words: anchor-free instance segmentation, Deep Learning(DL), encoder-decoder structure, attention mechanism, dilated convolution

中图分类号:

TP391

刘腾, 刘宏哲, 李学伟, 徐成. 基于无锚框分割网络改进的实例分割方法[J]. 计算机工程, 2022, 48(9): 239-247,253.

LIU Teng, LIU Hongzhe, LI Xuewei, XU Cheng. Improved Instance Segmentation Method Based on Anchor-Free Segmentation Network[J]. Computer Engineering, 2022, 48(9): 239-247,253.

http://www.ecice06.com/CN/Y2022/V48/I9/239

图/表 17

20221017110754

20221017110757

20221017110801

20221017110804

20221017110808

20221017110813

20221017110817

20221017110821

20221017110824

20221017110830

20221017110834

20221017110838

20221017110841

20221017110845

20221017110849

20221017110856

20221017110900

参考文献

[1] HE K M, GKIOXARI G, DOLLÁR P, et al.Mask R-CNN[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2017:2961-2969.
[2] REDMON J, DIVVALA S, GIRSHICK R, et al.You only look once:unified, real-time object detection[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2016:779-788.
[3] GIRSHICK R, DONAHUE J, DARRELL T, et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2014:580-587.
[4] TIAN Z, SHEN C H, CHEN H, et al.FCOS:fully convolutional one-stage object detection[C]//Proceedings of IEEE/CVF International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2019:9626-9635.
[5] ZHOU X Y, WANG D Q, KRÄHENBÜHL P.Objects as points[EB/OL].[2021-08-25].https://arxiv.org/abs/1904.07850.
[6] GIRSHICK R.Fast R-CNN[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2015:1440-1448.
[7] 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.
[8] REDMON J, FARHADI A.YOLO9000:better, faster, stronger[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2017:6517-6525.
[9] REDMON J, FARHADI A.YOLOv3:an incremental improvement[EB/OL].[2021-08-25].https://arxiv.org/pdf/1804.02767.pdf.
[10] LIU W, ANGUELOV D, ERHAN D, et al.SSD:single shot MultiBox detector[C]//Proceedings of European Conference on Computer Vision.Berlin, Germany:Springer, 2016:21-37.
[11] 高宗, 李少波, 陈济楠, 等.基于YOLO网络的行人检测方法[J].计算机工程, 2018, 44(5):215-219, 226. GAO Z, LI S B, CHEN J N, et al.Pedestrian detection method based on YOLO network[J].Computer Engineering, 2018, 44(5):215-219, 226.(in Chinese)
[12] NEUBECK A, VAN GOOL L.Efficient non-maximum suppression[C]//Proceedings of the 18th International Conference on Pattern Recognition.Washington D.C., USA:IEEE Press, 2006:850-855.
[13] LI Y, QI H Z, DAI J F, et al.Fully convolutional instance-aware semantic segmentation[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2017:4438-4446.
[14] CHEN H, SUN K Y, TIAN Z, et al.BlendMask:top-down meets bottom-up for instance segmentation[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2020:8570-8578.
[15] 伍锡如, 邱涛涛, 王耀南.改进Mask R-CNN的交通场景多目标快速检测与分割[J].仪器仪表学报, 2021, 42(7):242-249. WU X R, QIU T T, WANG Y N.Multi-object detection and segmentation for traffic scene based on improved Mask R-CNN[J].Chinese Journal of Scientific Instrument, 2021, 42(7):242-249.(in Chinese)
[16] 闫天冉, 马晓静, 饶颖露, 等.基于改进Mask R-CNN的建筑钢筋尺寸检测算法[J].计算机工程, 2021, 47(9):274-281. YAN T R, MA X J, RAO Y L, et al.Rebar size detection algorithm for intelligent construction supervision based on improved Mask R-CNN[J].Computer Engineering, 2021, 47(9):274-281.(in Chinese)
[17] BOLYA D, ZHOU C, XIAO F Y, et al.YOLACT:real-time instance segmentation[C]//Proceedings of IEEE/CVF International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2019:9156-9165.
[18] WANG X L, KONG T, SHEN C H, et al.SOLO:segmenting objects by locations[C]//Proceedings of European Conference on Computer Vision.Berlin, Germany:Springer, 2020:649-665.
[19] CHEN X L, GIRSHICK R, HE K M, et al.TensorMask:a foundation for dense object segmentation[C]//Proceedings of IEEE/CVF International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2019:2061-2069.
[20] XIE E Z, SUN P Z, SONG X G, et al.PolarMask:single shot instance segmentation with polar representation[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2020:12190-12199.
[21] XIE E Z, WANG W H, DING M Y, et al.PolarMask++:enhanced polar representation for single-shot instance segmentation and beyond[EB/OL].[2021-08-25].https://arxiv.org/abs/2105.02184.
[22] 马子彤, 王国栋.基于双流卷积神经网络的人体实例分割[J/OL].激光与光电子学进展:1-12[2021-09-08].http://kns.cnki.net/kcms/detail/31.1690.tn.20210712.1628.030.html. MA Z T, WANG G D.Human instance segmentation based on two-stream convolutional neural network[J/OL].Laser & Optoelectronics Progress:1-12[2021-09-08].http://kns.cnki.net/kcms/detail/31.1690.tn.20210712.1628.030.html. (in Chinese)
[23] 刘文波, 叶涛, 李颀.基于改进SOLO v2的番茄叶部病害检测方法[J].农业机械学报, 2021, 52(8):213-220. LIU W B, YE T, LI Q.Tomato leaf disease detection method based on improved SOLO v2[J].Transactions of the Chinese Society for Agricultural Machinery, 2021, 52(8):213-220.(in Chinese)
[24] 史彩娟, 陈厚儒, 葛录录, 等.注意力残差多尺度特征增强的显著性实例分割[J].图学学报, 2021, 42(6):883-890. SHI C J, CHEN H R, GE L L, et al.Salient instance segmentation via attention residual multi-scale feature enhancement[J].Journal of Graphics, 2021, 42(6):883-890.(in Chinese)
[25] CHEN L C, ZHU Y K, PAPANDREOU G, et al.Encoder-Decoder with atrous separable convolution for semantic image segmentation[C]//Proceedings of European Conference on Computer Vision.Berlin, Germany:Springer, 2018:833-851.
[26] DAI J F, QI H Z, XIONG Y W, et al.Deformable convolutional networks[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2017:764-773.
[27] YU F, KOLTUN V.Multi-scale context aggregation by dilated convolutions[EB/OL].[2021-08-25].https://arxiv.org/pdf/1511.07122.pdf.
[28] HUANG J J, ZHU Z, HUANG G.Multi-stage HRNet:multiple stage high-resolution network for human pose estimation[EB/OL].[2021-08-25].https://arxiv.org/abs/1910.05901.
[29] 王冲, 赵志刚, 潘振宽, 等.基于空洞卷积的人体实例分割算法[J].青岛大学学报(自然科学版), 2021, 34(2):53-58. WANG C, ZHAO Z G, PAN Z K, et al.Person instance segmentation algorithm based on dilated convolution[J].Journal of Qingdao University (Natural Science Edition), 2021, 34(2):53-58.(in Chinese)

选择文件类型/文献管理软件名称

选择包含的内容