Self-distillation object segmentation method via scale-attention knowledge transfer

doi:10.11772/j.issn.1001-9081.2023010075

Journal of Computer Applications ›› 2024, Vol. 44 ›› Issue (1): 129-137.DOI: 10.11772/j.issn.1001-9081.2023010075

• Artificial intelligence • Previous Articles

Self-distillation object segmentation method via scale-attention knowledge transfer

Xiaobing WANG¹^,², Xiongwei ZHANG¹(), Tieyong CAO¹, Yunfei ZHENG¹^,²^,³, Yong WANG²^,³

^1.Institute of Command and Control Engineering，Army Engineering University，Nanjing Jiangsu 210001，China
^2.Army Academy of Artillery and Air Defense （Nanjing Campus），Nanjing Jiangsu 211131，China
^3.Anhui Key Laboratory of Polarization Imaging and Detection（Army Academy of Artillery and Air Defense），Hefei Anhui 230031，China

Received:2023-01-31 Revised:2023-04-25 Accepted:2023-05-04 Online:2023-06-06 Published:2024-01-10
Contact: Xiongwei ZHANG
About author:WANG Xiaobing， born in 1981， Ph. D.， lecturer. His research interests include intelligent information processing， deep learning.
CAO Tieyong， born in 1971， Ph. D.， professor. His research interests include image processing， machine learning.
ZHENG Yunfei， born in 1983， Ph. D.， lecturer. His research interests include camouflage object recognition， deep learning.
WANG Yong， born in 1983， M. S.， lecturer. His research interests include multimodal image processing， pattern recognition.
Supported by:
National Natural Science Foundation of China(61801512)

基于尺度注意知识迁移的自蒸馏目标分割方法

王晓兵¹^,², 张雄伟¹(), 曹铁勇¹, 郑云飞¹^,²^,³, 王勇²^,³

^1.陆军工程大学指挥控制工程学院, 南京 210001
^2.陆军炮兵防空兵学院(南京校区), 南京 211131
^3.安徽省偏振成像与探测重点实验室(陆军炮兵防空兵学院), 合肥 230031

通讯作者: 张雄伟
作者简介:王晓兵（1981—），男，安徽滁州人，讲师，博士，主要研究方向：智能信息处理、深度学习；
曹铁勇（1971—），男，江苏南京人，教授，博士，主要研究方向：图像处理、机器学习；
郑云飞（1983—），男，安徽滁州人，讲师，博士，主要研究方向：伪装目标识别、深度学习；
王勇（1983—），男，安徽合肥人，讲师，硕士，主要研究方向：多模态图像处理、模式识别。
第一联系人：张雄伟（1965—），男，浙江嘉兴人，教授，博士，主要研究方向：多媒体信息处理、机器学习；
基金资助:
国家自然科学基金资助项目(61801512)

Abstract

Abstract:

It is difficult for current object segmentation models to reach a good balance between segmentation performance and inference efficiency. To solve this challenge， a self-distillation object segmentation method via scale-attention knowledge transfer was proposed. Firstly， an object segmentation network only using features in backbone was constructed as the inference network， to achieve efficient forward inference process. Secondly， a self-distillation learning model via scale-attention knowledge was proposed. On the one hand， a scale-attention pyramid feature module was designed to adaptively capture context information at different semantic levels and extract more discriminative self-distillation knowledge. On the other hand， a distillation loss was constructed by fusing cross entropy， KL （Kullback-Leibler） divergence and L2 distance. It drove distillation knowledge to transfer into segmentation network efficiently to improve its generalization performance. The method was verified on five public object segmentation datasets of COD （Camouflaged Object Detection）， DUT-O （Dalian University of Technology-OMRON）， SOC （Salient Objects in Clutter）， etc.： considering the proposed inference network as the baseline network， the proposed self-distillation model can increase the segmentation performance by 3.01% on F_β metric， which was 1.00% higher better than that of Teacher-Free （TF） self-distillation model； compared with recent Residual learning Net （R2Net）， the proposed object segmentation network reduces the number of parameters by 2.33×106， improves the inference frame rate by 2.53%， decreases the floating-point operations by 40.50%， and increases segmentation performance by 0.51%. Experimental results show that the proposed self-distillation segmentation method can balance performance and efficiency， and is suitable for scenarios with limited computing and storage resources.

Key words: self-distillation, object segmentation, knowledge transfer, scale-attention mechanism, pyramid knowledge representation

摘要：

当前的目标分割模型难以兼顾分割性能与推断效率，为此提出一种基于尺度注意知识迁移的自蒸馏目标分割方法。首先，构建了一个仅利用主干特征的目标分割网络作为推断网络，实现高效的前向推断过程。其次，提出了一种基于尺度注意知识的自蒸馏学习模型：一方面，设计了具有尺度注意机制的金字塔特征模块，利用尺度注意机制自适应地捕获不同语义水平的上下文信息，提取更具区分性的自蒸馏知识；另一方面，融合交叉熵、KL（Kullback-Leibler）散度和L2距离构造蒸馏损失，高效驱动蒸馏知识向分割网络迁移，提升泛化性能。该方法在COD（Camouflaged Object Detection）、DUT-O（Dalian University of Technology-OMRON）、SOC（Salient Objects in Clutter）等五个目标分割数据集上进行了验证：将所提推断网络作为基准网络，所提自蒸馏模型分割性能在F_β 指标上平均提升3.01%，比免教师（TF）自蒸馏模型增加了1.00%；所提网络与近期的残差分割网络（R2Net）相比，参数量减少了2.33×10⁶，推断帧率提升了2.53%，浮点运算量减少了40.50%，分割性能提升了0.51%。实验结果表明：所提方法能有效兼顾性能与效率，适用于计算和存储资源受限的应用场景。

关键词: 自蒸馏, 目标分割, 知识迁移, 尺度注意机制, 金字塔知识表示

CLC Number:

TP391.41

Xiaobing WANG, Xiongwei ZHANG, Tieyong CAO, Yunfei ZHENG, Yong WANG. Self-distillation object segmentation method via scale-attention knowledge transfer[J]. Journal of Computer Applications, 2024, 44(1): 129-137.

王晓兵, 张雄伟, 曹铁勇, 郑云飞, 王勇. 基于尺度注意知识迁移的自蒸馏目标分割方法[J]. 《计算机应用》唯一官方网站, 2024, 44(1): 129-137.

Figures/Tables 9

References 41

1	HE K， ZHANG X， REN S， et al. Deep residual learning for image recognition ［C］// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2016： 770-778. 10.1109/cvpr.2016.90
2	WANG J， SUN K， CHENG T， et al. Deep high-resolution representation learning for visual recognition ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2021， 43（10）： 3349-3364. 10.1109/tpami.2020.2983686
3	HUANG G， LIU Z， VAN DER MAATEN L， et al. Densely connected convolutional networks ［C］// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2017： 2261-2269. 10.1109/cvpr.2017.243
4	HINTON G， VINYALS O， DEAN J. Distilling the knowledge in a neural network ［EB/OL］. （2015-03-09）［2020-11-10］. .
5	LUO J H， WU J， LIN W. ThiNet： A filter level pruning method for deep neural network compression ［C］// Proceedings of the 2017 IEEE International Conference on Computer Vision. Piscataway： IEEE， 2017： 5068-5076. 10.1109/iccv.2017.541
6	RASTEGARI M， ORDONEZ V， REDMONON J， et al. XNOR-Net： ImageNet classification using binary convolutional neural networks ［C］// Proceedings of the 2016 European Conference on Computer Vision， LNCS 9908. Cham： Springer， 2016： 525-542.
7	FURLANELLO T， LIPTON Z C， TSCHANNEN M， et al. Born again neural networks ［C］// Proceedings of the 35th International Conference on Machine Learning. New York： JMLR.org， 2018： 1607-1616.
8	SUN D， YAO A， ZHOU A， et al. Deeply-supervised knowledge synergy ［C］// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2019： 6990-6999. 10.1109/cvpr.2019.00716
9	HOU Y， MA Z， LIU C， et al. Learning lightweight lane detection CNNs by self attention distillation ［C］// Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Piscataway： IEEE， 2019： 1013-1021. 10.1109/iccv.2019.00110
10	LI D， CHEN Q. Dynamic hierarchical mimicking towards consistent optimization objectives ［C］// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2020： 7639-7648. 10.1109/cvpr42600.2020.00766
11	JI M， SHIN S， HWANG S， et al. Refine myself by teaching myself： feature refinement via self-knowledge distillation ［C］// Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2021： 10659-10668. 10.1109/cvpr46437.2021.01052
12	ZHANG L， BAO C， MA K. Self-distillation： towards efficient and compact neural networks ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2022， 44（8）： 4388-4403.
13	郑云飞，王晓兵，张雄伟，等.基于金字塔知识的自蒸馏HRNet目标分割方法［J］.电子学报， 2023， 51（3）： 746-756. 10.12263/DZXB.20210169
	ZHENG Y F， WANG X B， ZHANG X W， et al. The self-distillation HRNet object segmentation based on the pyramid knowledge ［J］. Acta Electronica Sinica， 2023， 51（3）： 746-756. 10.12263/DZXB.20210169
14	YUAN L， TAY F E H， LI G， et al. Revisiting knowledge distillation via label smoothing regularization ［C］// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2020： 3902-3910. 10.1109/cvpr42600.2020.00396
15	YUN S， PARK J， LEE K， et al. Regularizing class-wise predictions via self-knowledge distillation ［C］// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2020： 13873-13882. 10.1109/cvpr42600.2020.01389
16	XU T B， LIU C L. Data-distortion guided self-distillation for deep neural networks ［C］// Proceedings of the 33rd AAAI Conference on Artificial Intelligence. Palo Alto， CA： AAAI Press， 2019： 5565-5572. 10.1609/aaai.v33i01.33015565
17	LEE H， HWANG S J， SHIN J. Self-supervised label augmentation via input transformations ［C］// Proceedings of the 2020 International Conference on Machine Learning. New York： JMLR.org， 2020： 5714-5724.
18	ZHOU B， BAU D， OLIVA A， et al. Interpreting deep visual representations via network dissection ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2019， 41（9）： 2131-2145. 10.1109/tpami.2018.2858759
19	HUANG Z， WANG X， HUANG L， et al. CCNet： criss-cross attention for semantic segmentation ［C］// Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Piscataway： IEEE， 2019： 603-612. 10.1109/iccv.2019.00069
20	LI X， ZHONG Z， WU J， et al. Expectation-maximization attention networks for semantic segmentation ［C］// Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Piscataway： IEEE， 2019： 9166-9175. 10.1109/iccv.2019.00926
21	WU Z， SU L， HUANG Q. Cascaded partial decoder for fast and accurate salient object detection ［C］// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2019： 3902-3911. 10.1109/cvpr.2019.00403
22	FAN D P， JI G P， SUN G， et al. Camouflaged object detection ［C］// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition， Piscataway： IEEE， 2020： 2774-2784. 10.1109/cvpr42600.2020.00285
23	WANG L， CHEN R， ZHU L， et al. Deep sub-region network for salient object detection ［J］. IEEE Transactions on Circuits and Systems for Video Technology， 2021， 31（2）： 728-741. 10.1109/tcsvt.2020.2988768
24	GAO S H， TAN Y Q， CHENG M M， et al. Highly efficient salient object detection with 100k parameters ［C］// Proceedings of the 2020 European Conference on Computer Vision， LNCS 12351. Cham： Springer， 2020： 702-721.
25	ZHAO X， PANG Y， ZHANG L， et al. Suppress and balance： a simple gated network for salient object detection ［C］// Proceedings of the 2020 European Conference on Computer Vision， LNCS 12347. Cham： Springer， 2020： 35-51.
26	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.
	WEN K， TANG W W， XIONG J C. Real-time segmentation algorithm based on attention mechanism and effective factorized convolution ［J］. Journal of Computer Applications， 2022， 42（9）： 2659-2666.
27	林荐壮，杨文忠，谭思翔，等.融合滤波增强和反转注意力网络用于息肉分割［J］.计算机应用， 2023， 43（1）： 265-272.
	LIN J Z， YANG W Z， TAN S X， et al. Fusing filter enhancement and reverse attention network for polyp segmentation ［J］. Journal of Computer Applications， 2023， 43（1）： 265-272.
28	ALLEN-ZHU Z， LI Y. Towards understanding ensemble， knowledge distillation and self-distillation in deep learning ［EB/OL］. （2021-07-03）［2022-12-12］. .
29	ZHAO H， SHI J， QI X， et al. Pyramid scene parsing network ［C］// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2017： 2830-2839. 10.1109/cvpr.2017.660
30	CHEN L C， PAPANDREOU G， SCHROFF F， et al. Rethinking atrous convolution for semantic image segmentation ［EB/OL］. （2017-12-05）［2022-12-12］. . 10.1007/978-3-030-01234-2_49
31	QIAO Z， YUAN X， ZHUANG C， et al. Attention pyramid module for scene recognition ［C］// Proceedings of the 25th International Conference on Pattern Recognition. Piscataway： IEEE， 2021： 7521-7528. 10.1109/icpr48806.2021.9412235
32	REBUFFI S A， FONG R， JI X， et al. There and back again： Revisiting backpropagation saliency methods ［C］// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2020： 8836-8845. 10.1109/cvpr42600.2020.00886
33	ZHENG Y， ZHANG X， WANG F， et al. Detection of people with camouflage pattern via dense deconvolution network ［J］. IEEE Signal Processing Letters， 2019， 26（1）： 29-33. 10.1109/lsp.2018.2825959
34	YANG C， ZHANG L， LU H， et al. Saliency detection via graph-based manifold ranking ［C］// Proceedings of the 2013 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2013： 3166-3173. 10.1109/cvpr.2013.407
35	FAN D P， CHENG M M， LIU J J， et al. Salient objects in clutter： Bringing salient object detection to the foreground ［C］// Proceedings of the 2018 European Conference on Computer Vision， LNCS 11219. Cham： Springer， 2018： 196-212.
36	CHENG M M， MITRA N J， HUANG X， et al. SalientShape： Group saliency in image collections ［J］. The Visual Computer， 2014， 30（4）： 443-453. 10.1007/s00371-013-0867-4
37	ACHANTA R， HEMAMI S， ESTRADA F， et al. Frequency-tuned salient region detection ［C］// Proceedings of the 2009 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2009： 1597-1604. 10.1109/cvpr.2009.5206596
38	DENG J， DONG W， SOCHER R， et al. ImageNet： A large-scale hierarchical image database ［C］// Proceedings of the 2009 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2009： 248-255. 10.1109/cvpr.2009.5206848
39	GAO S H， CHENG M M， ZHAO K， et al. Res2Net： A new multi-scale backbone architecture ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2021， 43（2）： 652-662. 10.1109/tpami.2019.2938758
40	FENG M， LU H， YU Y. Residual learning for salient object detection ［J］. IEEE Transactions on Image Processing， 2020， 29： 4696-4708. 10.1109/tip.2020.2975919
41	LIU J J， HOU Q， CHENG M M， et al. A simple pooling-based design for real-time salient object detection ［C］// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway： IEEE， 2019： 3912-3921. 10.1109/cvpr.2019.00404

网络	COD		CAMP		DUT-O		SOC		THUR		平均值
网络	F_β	MAE	F_β	MAE	F_β	MAE	F_β	MAE	F_β	MAE	F_β	MAE
ENet	63.07	6.36	59.23	0.95	78.38	5.39	86.83	5.93	82.60	4.97	74.02	4.72
POOL+R	61.55	6.50	58.29	1.29	82.95	4.41	87.92	4.77	85.25	3.78	75.19	4.15
SINet	66.54	6.42	69.73	1.73	82.76	4.90	85.41	10.00	86.81	3.63	78.25	5.34
R2Net	65.42	5.59	75.61	0.69	83.44	4.48	88.49	4.74	88.40	2.96	80.27	3.69
CCNet	64.44	4.78	58.29	0.76	79.70	4.68	87.27	4.72	84.80	3.57	74.90	3.70
CSFNet+R	58.66	6.94	35.77	1.82	77.96	6.02	85.89	5.82	84.93	3.68	68.64	4.86
GATENet	65.81	5.72	68.21	0.80	82.22	4.79	88.20	4.88	87.59	3.32	78.41	3.90
DSR	54.68	7.22	46.69	1.07	83.88	4.83	82.44	7.42	84.04	3.77	70.35	4.86
CPD	60.42	7.09	67.35	1.63	80.63	4.42	83.59	12.81	87.90	3.15	75.98	5.82
本文网络	68.10	4.54	74.28	0.51	83.61	3.98	89.01	4.39	88.40	2.65	80.68	3.21

网络	COD		CAMP		DUT-O		SOC		THUR		平均值
网络	F_β	MAE	F_β	MAE	F_β	MAE	F_β	MAE	F_β	MAE	F_β	MAE
ENet	63.07	6.36	59.23	0.95	78.38	5.39	86.83	5.93	82.60	4.97	74.02	4.72
POOL+R	61.55	6.50	58.29	1.29	82.95	4.41	87.92	4.77	85.25	3.78	75.19	4.15
SINet	66.54	6.42	69.73	1.73	82.76	4.90	85.41	10.00	86.81	3.63	78.25	5.34
R2Net	65.42	5.59	75.61	0.69	83.44	4.48	88.49	4.74	88.40	2.96	80.27	3.69
CCNet	64.44	4.78	58.29	0.76	79.70	4.68	87.27	4.72	84.80	3.57	74.90	3.70
CSFNet+R	58.66	6.94	35.77	1.82	77.96	6.02	85.89	5.82	84.93	3.68	68.64	4.86
GATENet	65.81	5.72	68.21	0.80	82.22	4.79	88.20	4.88	87.59	3.32	78.41	3.90
DSR	54.68	7.22	46.69	1.07	83.88	4.83	82.44	7.42	84.04	3.77	70.35	4.86
CPD	60.42	7.09	67.35	1.63	80.63	4.42	83.59	12.81	87.90	3.15	75.98	5.82
本文网络	68.10	4.54	74.28	0.51	83.61	3.98	89.01	4.39	88.40	2.65	80.68	3.21

网络	参数量/10⁶	推断帧率/（frame·s^-1）	浮点运算量/GFLOPs
ENet	34.80	37.59	29.19
POOL+R	70.50	21.53	38.46
SINet	48.90	30.10	7.86
R2Net	26.12	37.21	7.26
CCNet	52.10	35.34	61.57
CSFNet+R	36.50	36.68	6.04
GATENet	28.63	33.03	55.17
DSR	47.85	8.80	55.13
CPD	75.29	32.60	7.19
本文网络	23.79	38.15	4.32

网络	参数量/10⁶	推断帧率/（frame·s^-1）	浮点运算量/GFLOPs
ENet	34.80	37.59	29.19
POOL+R	70.50	21.53	38.46
SINet	48.90	30.10	7.86
R2Net	26.12	37.21	7.26
CCNet	52.10	35.34	61.57
CSFNet+R	36.50	36.68	6.04
GATENet	28.63	33.03	55.17
DSR	47.85	8.80	55.13
CPD	75.29	32.60	7.19
本文网络	23.79	38.15	4.32

模型	COD		CAMP		DUT-O		SOC		THUR		平均值
模型	F_β	MAE	F_β	MAE	F_β	MAE	F_β	MAE	F_β	MAE	F_β	MAE
基准网络	65.50	4.70	69.77	0.60	81.35	4.55	87.66	4.66	87.30	2.99	78.32	3.50
BL+DKS	66.90	4.54	71.75	0.57	82.18	4.27	88.18	4.91	87.62	2.81	79.48（+1.16）	3.42（-0.06）
BL+BYOT	66.89	4.57	71.83	0.56	82.26	4.22	87.98	4.44	87.72	2.89	79.33（+1.01）	3.34（-0.15）
BL+DHM	65.45	4.66	72.16	0.57	82.43	4.43	87.85	4.64	88.03	2.98	79.18（+0.86）	3.46（-0.04）
BL+SA	67.10	4.59	70.51	0.56	82.07	4.27	88.29	4.55	87.53	2.86	79.10（+0.78）	3.37（-0.13）
BL+TF	67.08	4.61	72.40	0.52	83.08	3.99	88.49	4.46	88.33	2.72	79.88（+1.56）	3.26（-0.24）
BL+FR	65.80	4.60	69.98	0.59	81.74	4.49	88.30	4.61	87.14	2.30	78.59（+0.27）	3.31（-0.18）
本文模型	68.10	4.54	74.28	0.51	83.61	3.98	89.01	4.39	88.40	2.65	80.68（+2.36）	3.21（-0.29）

Self-distillation object segmentation method via scale-attention knowledge transfer

基于尺度注意知识迁移的自蒸馏目标分割方法

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 9

References 41

Related Articles 15

Recommended Articles

Metrics

序号	推断网络损失度量	自蒸馏分支损失度量			知识匹配模式		平均值
序号	交叉熵	交叉熵	KL散度	L2	自上而下	一致性	F_β	MAE
1	√	—	—	—	—	—	81.35	4.55
2	√	√	—	—	—	—	82.36（+1.01）	4.18（-0.37）
3	√	√	√	—	√	—	83.04（+1.75）	4.04（-0.51）
4	√	√	√	—	—	√	82.67（+1.32）	4.13（-0.42）
5	√	√	√	—	√	√	83.44（+2.09）	4.01（-0.54）
6	√	√	√	√	√	—	83.23（+1.88）	4.03（-0.52）
7	√	√	√	√	—	√	83.14（+1.79）	4.07（-0.48）
8	√	√	√	√	√	√	83.61（+2.26）	3.98（-0.57）

[1]	Chunlan ZHAN, Anzhi WANG, Minghui WANG. Camouflage object segmentation method based on channel attention and edge fusion [J]. Journal of Computer Applications, 2023, 43(7): 2166-2172.
[2]	Yang LIU, Zhiyang LU, Jun WANG, Jun SHI. Gibbs artifact removal algorithm for magnetic resonance imaging based on self-attention connection UNet [J]. Journal of Computer Applications, 2023, 43(5): 1606-1611.
[3]	Zhongbo HU, Xupeng WANG. Multifactorial backtracking search optimization algorithm for solving automated test case generation problem [J]. Journal of Computer Applications, 2023, 43(4): 1214-1219.
[4]	Xiao LYU, Huihui SONG, Jiaqing FAN. Semi-supervised video object segmentation via deep and shallow representations fusion [J]. Journal of Computer Applications, 2022, 42(12): 3884-3890.
[5]	Caitong BAI, Xiaolong CUI, Huiji ZHENG, Ai LI. Robust speech recognition technology based on self-supervised knowledge transfer [J]. Journal of Computer Applications, 2022, 42(10): 3217-3223.
[6]	WANG Ning, SONG Huihui, ZHANG Kaihua. Accurate object tracking algorithm based on distance weighting overlap prediction and ellipse fitting optimization [J]. Journal of Computer Applications, 2021, 41(4): 1100-1105.
[7]	WEI Chunwu, ZHAO Juanjuan, TANG Xiaoxian, QIANG Yan. Knowledge extraction method for follow-up data based on multi-term distillation network [J]. Journal of Computer Applications, 2021, 41(10): 2871-2878.
[8]	JIANG Sihao, SONG Huihui, ZHANG Kaihua, TANG Runfa. Video object segmentation method based on dual pyramid network [J]. Journal of Computer Applications, 2019, 39(8): 2242-2246.
[9]	YU Huangyue, WANG Han, GUO Mengting. Video keyframe extraction based on users' interests [J]. Journal of Computer Applications, 2017, 37(11): 3139-3144.
[10]	. Moving object segmentation algorithm for river surveillance video [J]. Journal of Computer Applications, 2011, 31(01): 263-265.
[11]	. Video object segmentation in H.264 compressed domain based on entropy energy [J]. Journal of Computer Applications, 2010, 30(12): 3265-3268.
[12]	Sen Guo Wei Liu JiangHua Wang. Touched human object segmentation based on Mean-shift algorithm [J]. Journal of Computer Applications, 2009, 29(1): 51-53.
[13]	. Moving shadow detection algorithm based on region and illumination invariant [J]. Journal of Computer Applications, 2007, 27(9): 2152-2153.
[14]	LIU Zhen，ZHAO Jie-yu. Video segmentation method based on mixture probabilistic background model [J]. Journal of Computer Applications, 2005, 25(07): 1616-1619.
[15]	REN Jian-feng,GUO Lei,SHEN Yu-li. Motion segmentation based on multi-model robust estimation [J]. Journal of Computer Applications, 2005, 25(04): 778-780.