《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (9): 2659-2666.DOI: 10.11772/j.issn.1001-9081.2021071327
• 人工智能 • 上一篇
收稿日期:
2021-07-23
修回日期:
2021-10-14
接受日期:
2021-10-18
发布日期:
2021-10-29
出版日期:
2022-09-10
通讯作者:
唐伟伟
作者简介:
文凯(1972—),男,重庆人,高级工程师,博士,主要研究方向:大数据、计算机视觉、移动通信;Kai WEN1,2, Weiwei TANG1,2(), Junchen XIONG1,2
Received:
2021-07-23
Revised:
2021-10-14
Accepted:
2021-10-18
Online:
2021-10-29
Published:
2022-09-10
Contact:
Weiwei TANG
About author:
WEN Kai, born in 1972, Ph. D., senior engineer. His research interests include big data, computer vision, mobile communication.摘要:
针对现阶段实时语义分割算法计算成本高和内存占用大而无法满足实际场景需求的问题,提出一种新型的浅层的轻量级实时语义分割算法——基于注意力机制和有效分解卷积的实时分割算法(AEFNet)。首先,利用一维非瓶颈结构(Non-bottleneck-1D)构建轻量级分解卷积模块以提取丰富的上下文信息并减少运算量,同时以一种简单的方式增强算法学习能力并利于提取细节信息;然后,结合池化操作和注意力细化模块(ARM)构建全局上下文注意力模块以捕捉全局信息并细化算法的每个阶段,从而优化分割效果。算法在公共数据集cityscapes和camvid上进行验证,并在cityscapes测试集上获得精度为74.0%和推理速度为118.9帧速率(FPS),相比深度非对称瓶颈网络(DABNet),所提算法在精度上提高了约4个百分点,推理速度提升了14.7 FPS,与最近高效的增强非对称卷积网络(EACNet)相比,所提算法精度略低0.2个百分点,然而推理速度提高了6.9 FPS。实验结果表明:所提算法能够较为准确地识别场景信息,并能满足实时性要求。
中图分类号:
文凯, 唐伟伟, 熊俊臣. 基于注意力机制和有效分解卷积的实时分割算法[J]. 计算机应用, 2022, 42(9): 2659-2666.
Kai WEN, Weiwei TANG, Junchen XIONG. Real-time segmentation algorithm based on attention mechanism and effective factorized convolution[J]. Journal of Computer Applications, 2022, 42(9): 2659-2666.
模块 | mIOU/% | FPS |
---|---|---|
AEFNet | 74.0 | 118.9 |
AEFNet(r=4,4,4,4,4,4) | 72.7 | 118.7 |
AEFNet(r=3,3,7,7,13,13) | 73.3 | 118.2 |
AEFNet+ERFNetdecoder | 74.3 | 69.7 |
表1 空洞率和解码器对算法性能的影响
Tab.1 Influence of dilated rate and decoder on algorithm performance
模块 | mIOU/% | FPS |
---|---|---|
AEFNet | 74.0 | 118.9 |
AEFNet(r=4,4,4,4,4,4) | 72.7 | 118.7 |
AEFNet(r=3,3,7,7,13,13) | 73.3 | 118.2 |
AEFNet+ERFNetdecoder | 74.3 | 69.7 |
GCAM | mIOU/% | FPS | |
---|---|---|---|
Avg-pooling+AM | Max-pooling+AM | ||
√ | 73.1 | 119.5 | |
√ | 73.5 | 119.6 | |
72.6 | 125.6 | ||
√ | √ | 74.0 | 118.9 |
表2 GCAM对算法性能的影响
Tab.2 Influence of GCAM on algorithm performance
GCAM | mIOU/% | FPS | |
---|---|---|---|
Avg-pooling+AM | Max-pooling+AM | ||
√ | 73.1 | 119.5 | |
√ | 73.5 | 119.6 | |
72.6 | 125.6 | ||
√ | √ | 74.0 | 118.9 |
模块 | 长连接 | mIOU/% | FPS | ||
---|---|---|---|---|---|
GCAM | 阶段一 | 阶段二 | 阶段三 | ||
√ | √ | √ | 72.9 | 119.2 | |
√ | √ | √ | 73.6 | 117.3 | |
√ | √ | √ | 73.2 | 116.4 | |
√ | 73.8 | 119.3 |
表3 长连接对算法性能的影响
Tab. 3 Influence of long connection on algorithm performance
模块 | 长连接 | mIOU/% | FPS | ||
---|---|---|---|---|---|
GCAM | 阶段一 | 阶段二 | 阶段三 | ||
√ | √ | √ | 72.9 | 119.2 | |
√ | √ | √ | 73.6 | 117.3 | |
√ | √ | √ | 73.2 | 116.4 | |
√ | 73.8 | 119.3 |
算法 | 输入图片尺寸 | 预训练策略 | 参数量/MB | 精度/% | FPS |
---|---|---|---|---|---|
PSPNet[ | 713×713 | 有 | 65.70 | 78.4 | <1 |
SegNet[ | 360×640 | 有 | 29.50 | 56.1 | 14.6 |
ENet[ | 512×1024 | 无 | 0.36 | 58.3 | 76.9 |
ESPNet[ | 512×1024 | 无 | 0.36 | 60.3 | 112.0 |
ERFNet[ | 512×1024 | 无 | 2.10 | 68.0 | 41.7 |
ICNet[ | 1 024×2 048 | 有 | 7.80 | 69.5 | 30.3 |
AGLNet[ | 512×1 024 | 无 | 1.12 | 70.1 | 52.0 |
DABNet[ | 512×1 024 | 无 | 0.76 | 70.1 | 104.2 |
ESNet[ | 512×1 024 | 有 | 1.66 | 70.7 | 63.0 |
DFANet[ | 1 024×1 024 | 有 | 7.80 | 71.3 | 100.0 |
LRNNet[ | 512×1 024 | 无 | 0.68 | 72.2 | 71.0 |
EACNet[ | 512×1 024 | 无 | 1.10 | 74.2 | 113.0 |
AEFNet | 512×1 024 | 无 | 1.59 | 74.0 | 118.9 |
表4 不同算法在cityscapes测试集上精度与推理速度的对比
Tab.4 Precision and interference speed comparation of different algorithms on cityscapes test set
算法 | 输入图片尺寸 | 预训练策略 | 参数量/MB | 精度/% | FPS |
---|---|---|---|---|---|
PSPNet[ | 713×713 | 有 | 65.70 | 78.4 | <1 |
SegNet[ | 360×640 | 有 | 29.50 | 56.1 | 14.6 |
ENet[ | 512×1024 | 无 | 0.36 | 58.3 | 76.9 |
ESPNet[ | 512×1024 | 无 | 0.36 | 60.3 | 112.0 |
ERFNet[ | 512×1024 | 无 | 2.10 | 68.0 | 41.7 |
ICNet[ | 1 024×2 048 | 有 | 7.80 | 69.5 | 30.3 |
AGLNet[ | 512×1 024 | 无 | 1.12 | 70.1 | 52.0 |
DABNet[ | 512×1 024 | 无 | 0.76 | 70.1 | 104.2 |
ESNet[ | 512×1 024 | 有 | 1.66 | 70.7 | 63.0 |
DFANet[ | 1 024×1 024 | 有 | 7.80 | 71.3 | 100.0 |
LRNNet[ | 512×1 024 | 无 | 0.68 | 72.2 | 71.0 |
EACNet[ | 512×1 024 | 无 | 1.10 | 74.2 | 113.0 |
AEFNet | 512×1 024 | 无 | 1.59 | 74.0 | 118.9 |
测试 目标 | 算法模型 | |||||
---|---|---|---|---|---|---|
ENet | EFSNet | CGNet | ERFNet | DABNet | AEFNet | |
道路 | 96.3 | 96.6 | 95.5 | 97.9 | 96.8 | 98.0 |
人行道 | 74.2 | 74.9 | 78.7 | 82.1 | 78.5 | 84.4 |
建筑物 | 75.0 | 86.4 | 88.1 | 90.7 | 90.9 | 91.7 |
墙壁 | 32.2 | 37.5 | 40.0 | 45.2 | 45.3 | 48.8 |
栅栏 | 33.2 | 39.6 | 43.0 | 50.4 | 50.1 | 58.1 |
电杆 | 43.4 | 48.0 | 54.1 | 59.0 | 59.1 | 63.0 |
交通灯 | 34.1 | 49.8 | 59.8 | 62.6 | 65.2 | 67.7 |
交通标志 | 44.0 | 55.1 | 63.9 | 68.4 | 70.7 | 75.5 |
植物 | 88.6 | 89.7 | 89.6 | 91.9 | 92.5 | 92.3 |
地势 | 61.4 | 64.9 | 67.6 | 69.4 | 68.1 | 68.9 |
天空 | 90.6 | 92.8 | 92.9 | 94.2 | 94.6 | 93.9 |
行人 | 65.5 | 70.3 | 74.9 | 78.5 | 80.5 | 80.6 |
骑手 | 38.4 | 51.5 | 54.9 | 59.8 | 58.5 | 61.4 |
汽车 | 90.6 | 90.2 | 90.2 | 93.4 | 92.7 | 93.9 |
卡车 | 36.9 | 43.0 | 44.1 | 52.3 | 52.7 | 65.4 |
公交车 | 50.5 | 49.7 | 59.5 | 60.8 | 67.2 | 78.1 |
拖车 | 48.1 | 41.6 | 25.2 | 53.7 | 50.9 | 52.7 |
摩托车 | 38.8 | 41.5 | 47.3 | 49.9 | 50.4 | 57.6 |
自行车 | 55.4 | 53.5 | 60.2 | 64.2 | 65.7 | 74.4 |
类mIOU | 58.3 | 61.9 | 64.8 | 69.7 | 70.1 | 74.0 |
类别mIOU | 80.4 | 82.8 | 85.7 | 87.3 | 87.8 | 88.6 |
表5 AEFNet在cityscapes测试集上的每类IOU及与其他算法的比较 (%)
Tab.5 Each class IOU of AEFNet and other algorithms on cityscapes test set
测试 目标 | 算法模型 | |||||
---|---|---|---|---|---|---|
ENet | EFSNet | CGNet | ERFNet | DABNet | AEFNet | |
道路 | 96.3 | 96.6 | 95.5 | 97.9 | 96.8 | 98.0 |
人行道 | 74.2 | 74.9 | 78.7 | 82.1 | 78.5 | 84.4 |
建筑物 | 75.0 | 86.4 | 88.1 | 90.7 | 90.9 | 91.7 |
墙壁 | 32.2 | 37.5 | 40.0 | 45.2 | 45.3 | 48.8 |
栅栏 | 33.2 | 39.6 | 43.0 | 50.4 | 50.1 | 58.1 |
电杆 | 43.4 | 48.0 | 54.1 | 59.0 | 59.1 | 63.0 |
交通灯 | 34.1 | 49.8 | 59.8 | 62.6 | 65.2 | 67.7 |
交通标志 | 44.0 | 55.1 | 63.9 | 68.4 | 70.7 | 75.5 |
植物 | 88.6 | 89.7 | 89.6 | 91.9 | 92.5 | 92.3 |
地势 | 61.4 | 64.9 | 67.6 | 69.4 | 68.1 | 68.9 |
天空 | 90.6 | 92.8 | 92.9 | 94.2 | 94.6 | 93.9 |
行人 | 65.5 | 70.3 | 74.9 | 78.5 | 80.5 | 80.6 |
骑手 | 38.4 | 51.5 | 54.9 | 59.8 | 58.5 | 61.4 |
汽车 | 90.6 | 90.2 | 90.2 | 93.4 | 92.7 | 93.9 |
卡车 | 36.9 | 43.0 | 44.1 | 52.3 | 52.7 | 65.4 |
公交车 | 50.5 | 49.7 | 59.5 | 60.8 | 67.2 | 78.1 |
拖车 | 48.1 | 41.6 | 25.2 | 53.7 | 50.9 | 52.7 |
摩托车 | 38.8 | 41.5 | 47.3 | 49.9 | 50.4 | 57.6 |
自行车 | 55.4 | 53.5 | 60.2 | 64.2 | 65.7 | 74.4 |
类mIOU | 58.3 | 61.9 | 64.8 | 69.7 | 70.1 | 74.0 |
类别mIOU | 80.4 | 82.8 | 85.7 | 87.3 | 87.8 | 88.6 |
算法 | 输入图片尺寸 | 精度/% | 帧速率/FPS | 参数量/MB |
---|---|---|---|---|
ENet[ | 360×480 | 51.3 | 61.0 | 0.36 |
Segnet[ | 360×480 | 55.6 | 16.7 | 29.50 |
ESPNet[ | 360×480 | 55.6 | 132.0 | 0.36 |
EDANet[ | 360×480 | 66.4 | 163.0 | 0.68 |
DABNet[ | 360×480 | 66.4 | 117.0 | 0.76 |
ICNet[ | 720×960 | 67.1 | 30.3 | 7.80 |
DFANet[ | 360×480 | 71.3 | 100.0 | 7.80 |
AEFNet | 360×480 | 67.6 | 123.6 | 1.59 |
表6 不同算法在camvid测试集上的性能对比
Tab.6 Performance comparation of different algorithms on camvid test set
算法 | 输入图片尺寸 | 精度/% | 帧速率/FPS | 参数量/MB |
---|---|---|---|---|
ENet[ | 360×480 | 51.3 | 61.0 | 0.36 |
Segnet[ | 360×480 | 55.6 | 16.7 | 29.50 |
ESPNet[ | 360×480 | 55.6 | 132.0 | 0.36 |
EDANet[ | 360×480 | 66.4 | 163.0 | 0.68 |
DABNet[ | 360×480 | 66.4 | 117.0 | 0.76 |
ICNet[ | 720×960 | 67.1 | 30.3 | 7.80 |
DFANet[ | 360×480 | 71.3 | 100.0 | 7.80 |
AEFNet | 360×480 | 67.6 | 123.6 | 1.59 |
1 | LONG J, SHELHAMER E, DARRELL T. Fully convolutional networks for semantic segmentation[C]// Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2015: 3431-3440. 10.1109/cvpr.2015.7298965 |
2 | BADRINARAYANAN V, KENDALL A, CIPOLLA R. SegNet: a deep convolutional encoder-decoder architecture for image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(12): 2481-2495. 10.1109/tpami.2016.2644615 |
3 | ZHAO H S, SHI J P, QI X J, et al. Pyramid scene parsing network[C]// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017: 6230-6239. 10.1109/cvpr.2017.660 |
4 | LIN G S, MILAN A, SHEN C H, et al. RefineNet: multi-path refinement networks for high-resolution semantic segmentation[C]// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017:5168-5177. 10.1109/cvpr.2017.549 |
5 | PASZKE A, CHAURASIA A, KIM S, et al. ENet: a deep neural network architecture for real-time semantic segmentation[EB/OL]. (2016-06-07) [2021-07-15].. 10.48550/arXiv.1606.02147 |
6 | EMARA T, MUNIM H E ABD EL, ABBAS H M. LiteSeg: a novel lightweight ConvNet for semantic segmentation[C]// Proceedings of 2019 Digital Image Computing: Techniques and Applications. Piscataway: IEEE, 2019: 1-7. 10.1109/dicta47822.2019.8945975 |
7 | LI G, KIM J. DABNet: depth-wise asymmetric bottleneck for real-time semantic segmentation[C]// Proceedings of the 2019 British Machine Vision Conference. Durham: BMVA Press, 2019: No.186. 10.1109/access.2020.2971760 |
8 | LI Y Q, LI X K, XIAO C J, et al. EACNet: enhanced asymmetric convolution for real-time semantic segmentation[J]. IEEE Signal Processing Letters, 2021, 28: 234-238. 10.1109/lsp.2021.3051845 |
9 | HOWARD A G, ZHU M L, CHEN B, et al. MobileNets: efficient convolutional neural networks for mobile vision applications[EB/OL]. (2017-04-17) [2021-08-15].. 10.48550/arXiv.1704.04861 |
10 | ROMERA E, ÁLVAREZ J M, BERGASA L M, et al. ERFNet: efficient residual factorized ConvNet for real-time semantic segmentation[J]. IEEE Transactions on Intelligent Transportation Systems, 2018, 19(1): 263-272. 10.1109/tits.2017.2750080 |
11 | CHEN L C, PAPANDREOU G, SCHROFF F, et al. Rethinking atrous convolution for semantic image segmentation[EB/OL]. (2017-12-05) [2021-08-06].. 10.1007/978-3-030-01234-2_49 |
12 | MEHTA S, RASTEGARI M, CASPI A, et al. ESPNet: efficient spatial pyramid of dilated convolutions for semantic segmentation[C]// Proceedings of the 2018 European Conference on Computer Vision, LNCS 11214. Cham: Springer, 2018: 561-580. |
13 | YU C Q, WANG J B, PENG C, et al. BiSeNet: bilateral segmentation network for real-time semantic segmentation[C]// Proceedings of the 2018 European Conference on Computer Vision, LNCS 11217. Cham: Springer, 2018: 334-349. |
14 | FU J, LIU J, TIAN H J, et al. Dual attention network for scene segmentation[C]// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 3141-3149. 10.1109/cvpr.2019.00326 |
15 | HU P, PERAZZI F, HEILBRON F C, et al. Real-time semantic segmentation with fast attention[J]. IEEE Robotics and Automation Letters, 2021, 6(1): 263-270. 10.1109/lra.2020.3039744 |
16 | ZHOU W J, YUAN J Z, LEI J S, et al. TSNet: three-stream self-attention network for RGB-D indoor semantic segmentation[J]. IEEE Intelligent Systems, 2021, 36(4): 73-78. 10.1109/mis.2020.2999462 |
17 | HAN K, WANG Y H, TIAN Q, et al. GhostNet: more features from cheap operations[C]// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2020: 1577-1586. 10.1109/cvpr42600.2020.00165 |
18 | ZHAO H S, QI X J, SHEN X Y, et al. ICNet for real-time semantic segmentation on high-resolution images[C]// Proceedings of the 2018 European Conference on Computer Vision, LNCS 11207. Cham: Springer, 2018: 418-434. |
19 | ZHOU Q, WANG Y, FAN Y W, et al. AGLNet: towards real-time semantic segmentation of self-driving images via attention-guided lightweight network[J]. Applied Soft Computing, 2020, 96: No.106682. 10.1016/j.asoc.2020.106682 |
20 | WANG Y, ZHOU Q, XIONG J, et al. ESNet: an efficient symmetric network for real-time semantic segmentation[C]// Proceedings of the 2019 Chinese Conference on Pattern Recognition and Computer Vision, LNCS 11858. Cham: Springer, 2019: 41-52. |
21 | LI H C, XIONG P F, FAN H Q, et al. DFANet: deep feature aggregation for real-time semantic segmentation[C]// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 9514-9523. 10.1109/cvpr.2019.00975 |
22 | JIANG W H, XIE Z Z, LI Y Y, et al. LRNNet: a light-weighted network with efficient reduced non-local operation for real-time semantic segmentation[C]// Proceedings of 2020 IEEE International Conference on Multimedia and Expo Workshops. Piscataway: IEEE, 2020: 1-6. 10.1109/icmew46912.2020.9106038 |
23 | HU X G, WANG H B. Efficient fast semantic segmentation using continuous shuffle dilated convolutions[J]. IEEE Access, 2020, 8:70913-70924. 10.1109/access.2020.2987080 |
24 | WU T Y, TANG S, ZHANG R, et al. CGNet: a light-weight context guided network for semantic segmentation[J]. IEEE Transactions on Image Processing, 2021, 30:1169-1179. 10.1109/tip.2020.3042065 |
25 | LO S Y, HANG H M, CHAN S W, et al. Efficient dense modules of asymmetric convolution for real-time semantic segmentation[C]// Proceedings of the 2019 ACM International Conference on Multimedia in Asia. New York: ACM, 2019: No.1. 10.1145/3338533.3366558 |
26 | 高世伟,张长柱,王祝萍. 基于可分离金字塔的轻量级实时语义分割算法[J]. 计算机应用, 2021, 41(10): 2937-2944. 10.11772/j.issn.1001-9081.2020121939 |
GAO S W, ZHANG C Z, WANG Z P. Lightweight real-time semantic segmentation algorithm based on separable pyramid[J]. Journal of Computer Applications, 2021, 41(10): 2937-2944. 10.11772/j.issn.1001-9081.2020121939 | |
27 | 秦飞巍,沈希乐,彭勇,等. 无人驾驶中的场景实时语义分割方法[J]. 计算机辅助设计与图形学学报, 2021, 33(7):1026-1037. 10.3724/SP.J.1089.2021.18631 |
QIN F W, SHEN X Y, PENG Y, et al. A real-time semantic segmentation approach for autonomous driving scenes[J]. Journal of Computer-Aided Design and Graphics, 2021, 33(7): 1026-1037. 10.3724/SP.J.1089.2021.18631 | |
28 | 胡嵽,冯子亮. 基于深度学习的轻量级道路图像语义分割算法[J]. 计算机应用, 2021, 41(5):1326-1331. 10.11772/j.issn.1001-9081.2020081181 |
HU D, FENG Z L. Light-weight road image semantic segmentation algorithm based on deep learning[J]. Journal of Computer Applications, 2021, 41(5): 1326-1331. 10.11772/j.issn.1001-9081.2020081181 |
[1] | 张丽莹, 庞春江, 王新颖, 李国亮. 基于改进YOLOv3的多尺度目标检测算法[J]. 《计算机应用》唯一官方网站, 2022, 42(8): 2423-2431. |
[2] | 张新宇, 丁胜, 杨治佩. 基于改进注意力机制的交通标志检测算法[J]. 《计算机应用》唯一官方网站, 2022, 42(8): 2378-2385. |
[3] | 玄英律, 万源, 陈嘉慧. 基于多尺度卷积和注意力机制的LSTM时间序列分类[J]. 《计算机应用》唯一官方网站, 2022, 42(8): 2343-2352. |
[4] | 吴明晖, 张广洁, 金苍宏. 基于多模态信息融合的时间序列预测模型[J]. 《计算机应用》唯一官方网站, 2022, 42(8): 2326-2332. |
[5] | 吕振虎, 许新征, 张芳艳. 基于挤压激励的轻量化注意力机制模块[J]. 《计算机应用》唯一官方网站, 2022, 42(8): 2353-2360. |
[6] | 李坤, 侯庆. 基于注意力机制的轻量型人体姿态估计[J]. 《计算机应用》唯一官方网站, 2022, 42(8): 2407-2414. |
[7] | 徐成霞, 阎庆, 李腾, 苗开超. 基于联合注意力机制的单幅图像去雨算法[J]. 《计算机应用》唯一官方网站, 2022, 42(8): 2578-2585. |
[8] | 黄诚, 赵倩锐. 基于语言模型词嵌入和注意力机制的敏感信息检测方法[J]. 《计算机应用》唯一官方网站, 2022, 42(7): 2009-2014. |
[9] | 李晓寒, 王俊, 贾华丁, 萧刘. 基于多重注意力机制的图神经网络股市波动预测方法[J]. 《计算机应用》唯一官方网站, 2022, 42(7): 2265-2273. |
[10] | 左亚尧, 陈皓宇, 陈致然, 洪嘉伟, 陈坤. 融合多语义特征的命名实体识别方法[J]. 《计算机应用》唯一官方网站, 2022, 42(7): 2001-2008. |
[11] | 韩泽芳, 张雄, 上官宏, 韩兴隆, 韩静, 奉刚, 崔学英. 用于低剂量CT降噪的伪影感知生成对抗网络[J]. 《计算机应用》唯一官方网站, 2022, 42(7): 2301-2310. |
[12] | 秦庭威, 赵鹏程, 秦品乐, 曾建朝, 柴锐, 黄永琦. 基于残差注意力机制的点云配准算法[J]. 《计算机应用》唯一官方网站, 2022, 42(7): 2184-2191. |
[13] | 刘万军, 王佳铭, 曲海成, 董利兵, 曹欣宇. 基于频谱空间域特征注意的音乐流派分类算法[J]. 《计算机应用》唯一官方网站, 2022, 42(7): 2072-2077. |
[14] | 凡文俊, 赵曙光, 郭力争. 基于改进RetinaNet的船舶检测算法[J]. 《计算机应用》唯一官方网站, 2022, 42(7): 2248-2255. |
[15] | 刘博, 卿粼波, 王正勇, 刘美, 姜雪. 基于分块注意力机制和交互位置关系的群组活动识别[J]. 《计算机应用》唯一官方网站, 2022, 42(7): 2052-2057. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||