利用多种子信息聚合和正负混合学习的弱监督图像语义分割

桑雨; 刘通; 马天娇; 李乐; 李思漫; 刘宇男

doi:10.11999/JEIT250112

利用多种子信息聚合和正负混合学习的弱监督图像语义分割

doi: 10.11999/JEIT250112 cstr: 32379.14.JEIT250112

桑雨^{1, 2, ,},
刘通¹,
马天娇¹,
李乐¹,
李思漫¹,
刘宇男³

1.
辽宁工程技术大学电子与信息工程学院葫芦岛 125105
2.
渤海大学数学科学学院锦州 121013
3.
大连海事大学人工智能学院大连 116026

基金项目: 国家自然科学基金资助项目(62372077, 62302249)，中国博士后科学基金资助项目(2022M720624)，辽宁省教育厅科研基金(LJKQZ2021152)，国家重点研发计划项目(2019YFB2102400)，高校人才引进基金项目(18-1021)

详细信息

作者简介:
桑雨：男，副教授，研究方向为人工智能、视觉感知

刘通：女，硕士生，研究方向为人工智能、弱监督图像分割

马天娇：女，硕士，研究方向为人工智能、弱监督图像分割

李乐：男，硕士生，研究方向为深度学习、图像语义分割

李思漫：女，硕士生，研究方向为深度学习、目标检测

刘宇男：男，博士后研究员，研究方向为模式识别、计算机视觉

通讯作者:
桑雨　sangyu@lntu.edu.cn

中图分类号: TN911.73; TP391.41
计量
- 文章访问数: 72
- HTML全文浏览量: 45
- PDF下载量: 7
- 被引次数: 0
出版历程
- 收稿日期: 2025-02-26
- 修回日期: 2025-08-20
- 网络出版日期: 2025-08-27

Funds: National Natural Science Foundation of China (62372077, 62302249), China Postdoctoral Science Foundation (2022M720624), Research Fund of Liaoning Provincial Education Department (LJKQZ2021152), National Key Research and Development Program of China (2019YFB2102400), University Talent Introduction Foundation(18-1021)

摘要

摘要: 基于图像级标签的弱监督语义分割(WSSS)旨在通过类激活映射(CAM)生成伪标签(种子)，然后将其用于训练语义分割模型，为耗时且昂贵的像素级标注节省大量人力和财力。现有方法主要围绕CAM进行改进以获取单个优良的种子，同时通过一些后处理手段进一步提升种子的质量，但其得到的种子仍存在不等程度的噪声。为了减少噪声标签对分割网络造成的影响，考虑多个不同种子更能有效提取出正确信息，该文从多种子信息互补的角度，提出一种基于多种子信息聚合和正负混合学习的弱监督图像语义分割方法，通过在分类网络中改变输入图像尺度以及调整Dropout层随机隐藏神经元的概率，获取多个优良种子；依据它们对每个像素分配的类别标签情况进行优选获得聚合种子，并进一步区分该像素标签为干净标签还是噪声标签；利用正负混合学习训练语义分割网络，引入预测约束损失以避免网络对噪声标签给予过高的预测值，进而对干净标签应用正学习发挥正确信息的准确性，对噪声标签应用负学习抑制错误信息的影响，从而有效提升分割网络的性能。在PASCAL VOC 2012和MS COCO 2014验证集上实验结果表明，该文方法在基于卷积神经网络框架的分割网络中，mIoU分别达到了72.5%和40.8%，与RCA及URN方法相比分别提升了0.3%与0.1%；在基于Transformer框架的分割网络中，mIoU则提升至76.8%和46.7%，与CoBra及ECA方法相比分别提升了2.5%与1.6%，验证了方法的有效性。
- 弱监督语义分割 /
- 类激活映射 /
- 多种子 /
- 交叉熵 /
- 正负学习

HTML全文

图 1 本文方法整体框架图

下载: 全尺寸图片幻灯片

图 2 单种子与真值标签对比图

下载: 全尺寸图片幻灯片

图 3 多种子生成流程图

下载: 全尺寸图片幻灯片

图 4 生成6个种子示例图

下载: 全尺寸图片幻灯片

图 5 标签区分示例图

下载: 全尺寸图片幻灯片

图 6 聚合种子在不同阈值$t$下的区分结果，其中，绿色表示干净标签，红色表示噪声标签

下载: 全尺寸图片幻灯片

图 7 正负混合学习流程图

下载: 全尺寸图片幻灯片

图 8 在VOC验证集上的可视化分割结果

下载: 全尺寸图片幻灯片

图 9 在COCO验证集上的可视化分割结果

下载: 全尺寸图片幻灯片

表 1 本文方法与不同模型在PASCAL VOC 2012验证集与测试集上的性能对比(粗体表示最优值)

模型		骨干网络	mIoU(%)
模型		骨干网络	验证集	测试集
基于CNN方法	SIPE_(CVPR22)^[39]	Deeplab-V2+Resnet101	68.8	69.7
	FPR_(CVPR23)^[40]	Deeplab-V2+Resnet101	70.3	70.1
	ReCAM_(CVPR22)^[33]	Deeplab-V2+Resnet101	71.6	71.4
	EPS_(CVPR21)^[12]	Deeplab-V1+ResNet101	71.0	71.8
	BECO_(CVPR23)^[17]	Deeplab-V3+^[41]+Resnet101	72.1	71.8
	RCA_(CVPR23)^[34]	Deeplab-V2+ResNet101	72.2	72.8
	本文方法	Deeplab-V1+Resnet101	72.1	73.1
	本文方法	Deeplab-V2+Resnet101	72.5	73.9
基于Transformer方法	BECO_(CVPR23)^[17]	SegFormer-MiT-B2	73.7	73.5
	CoBra_(PR25)^[38]	SegFormer-MiT-B2	74.3	74.2
	本文方法	SegFormer-MiT-B2	75.3	75.9
	本文方法	Mask2Former+Swin-L	76.8	77.5

下载: 导出CSV

表 2 在PASCAL VOC 2012验证集上的每类分割结果(%，粗体表示最优值，下划线表示次优值)

方法	背景	飞机	自行车	鸟	船	瓶子	公共汽车	汽车	猫	椅子	牛	餐桌	狗	马	摩托车	人	盆栽植物	羊	沙发	火车	电视/ 显示器	mIoU
ADELE^[42]	91.1	77.6	33.0	88.9	67.1	71.7	88.8	82.5	89.0	26.6	83.8	44.6	84.4	77.8	74.8	78.5	43.8	84.8	44.6	56.1	65.3	69.3
W-OoD^[13]	91.0	80.1	34.1	88.1	64.8	68.3	87.4	84.4	89.8	30.1	87.8	34.7	87.5	85.9	79.8	75.0	56.4	84.5	47.8	80.4	46.4	70.7
EPS^[12]	91.7	89.4	40.6	84.7	67.0	71.6	87.8	82.7	87.4	33.6	81.9	37.3	82.5	82.9	76.6	82.8	54.0	79.7	39.1	85.4	51.7	71.0
EPS++^[43]	91.9	89.7	41.7	82.4	68.4	73.7	89.5	80.8	86.9	31.8	86.9	43.0	82.7	86.6	81.1	77.7	47.8	84.8	41.1	84.1	50.6	71.6
本文方法	92.2	89.5	40.1	89.5	74.8	76.2	87.6	82.2	88.3	31.9	85.8	34.7	85.8	82.3	74.2	82.0	55.6	86.2	36.8	83.7	53.9	72.1(V1)
本文方法	92.2	90.4	40.4	88.2	76.3	73.8	88.4	83.0	89.1	33.1	87.0	36.1	86.7	86.2	76.4	81.7	54.2	83.9	38.2	84.8	51.7	72.5(V2)

下载: 导出CSV

表 3 本文方法与不同模型在MS COCO 2014验证集上的性能对比(粗体表示最优值)

模型		骨干网络	mIoU(%)
模型		骨干网络	验证集
基于CNN 方法	EPS_(CVPR21)^[12]	Deeplab-V2+Resnet101	35.7
	OC-CSE_(ICCV21)^[8]	Deeplab-V1+Resnet38	36.4
	MDBA_(TIP23)^[16]	Deeplab-V2+Resnet101	37.8
	SIPE_(CVPR22)^[39]	Deeplab-V2+Resnet101	40.6
	URN_(AAAI22)^[44]	PSPNet+ResNet101	40.7
	本文方法	Deeplab-V2+Resnet101	40.8
基于 Transformer 方法	TSCD_(AAAI23)^[46]	SegFormer+MiT-B1	40.1
	ECA_{(ACM MM24)}^[45]	SegFormer+MiT-B1	42.9
	本文方法	SegFormer+MiT-B2	45.1
	本文方法	Mask2Former+Swin-L	46.7

下载: 导出CSV

表 4 在PASCAL VOC 2012训练集上种子的每类mIoU值(%，粗体表示最优值)

种子	背景	飞机	自行车	鸟	船	瓶子	公共汽车	汽车	猫	椅子	牛	餐桌	狗	马	摩托车	人	盆栽植物	羊	沙发	火车	电视/ 显示器	mIoU
0.05-1.0	91.4	87.9	45.9	85.0	80.6	75.3	83.7	81.0	86.0	34.9	83.5	44.4	85.4	86.1	80.8	80.0	54.7	86.7	41.0	80.1	54.9	72.8
0.05-1.2	91.4	87.3	44.5	87.0	80.8	73.5	83.6	80.7	86.8	34.5	85.1	48.7	83.3	87.0	79.9	80.7	52.5	85.9	41.9	78.9	55.4	72.8
0.10-1.0	91.5	87.3	45.0	86.5	80.2	75.2	84.3	82.0	85.7	34.5	84.0	43.6	84.7	86.0	80.8	80.3	57.3	89.3	40.6	79.0	55.9	73.0
0.10-1.2	91.5	88.0	41.9	86.2	80.1	74.0	82.9	80.6	87.7	34.6	84.9	49.6	84.6	86.5	79.9	80.7	54.0	87.3	41.9	78.6	56.5	73.0
0.20-1.0	91.5	87.0	45.9	85.9	80.2	74.1	83.6	80.7	86.6	35.3	83.7	46.0	85.1	86.3	80.3	80.0	55.4	86.8	38.4	80.1	55.5	72.8
0.20-1.2	91.5	87.8	45.1	86.7	79.4	72.5	84.1	82.0	87.1	34.2	84.5	47.6	83.9	86.8	79.8	80.8	55.0	86.6	41.1	79.1	55.8	72.9
0.30-1.0	91.3	87.5	41.2	85.5	79.6	73.8	83.7	81.4	86.3	34.0	84.3	44.8	85.4	86.8	80.5	79.5	54.7	86.7	40.4	78.0	54.4	72.4
0.30-1.2	91.5	87.3	44.4	86.6	78.9	72. 5	83.3	80.8	87.5	33.8	84.2	46.3	83.7	86.9	79.4	80.9	56.7	86.7	42.7	79.6	56.9	72.9

下载: 导出CSV

表 5 在PASCAL VOC 2012训练集上每个种子类激活图的mIoU值(%)

种子	0.05-1.0	0.05-1.2	0.10-1.0	0.10-1.2	0.2-1.0	0.3-1.2
mIoU	69.15	69.10	69.34	69.30	69.27	69.63

下载: 导出CSV

表 6 种子在PASCAL VOC 2012验证集中的mIoU值(%，粗体表示最优值)

Dropout	Scale	Deeplabv1+ResNet101	DeeplabV2+ResNet101
0.05	1.0	71.38	71.92
0.05	1.2	71.82	72.27
0.10	1.0	71.36	72.29
0.10	1.2	71.80	72.27
0.20	1.0	71.06	71.80
0.30	1.2	71.11	72.13
聚合种子	1.0	71.85	72.30

下载: 导出CSV

表 7 正负混合学习性能对比(%，粗体表示最优值)

学习方式		DeepLabV1+ResNet101	DeepLabV2+ResNet101
正负学习 (不同阈值t)	3	71.87	72.34
	4	71.89	72.37
	5	71.99	72.42
	6	72.06	72.46

下载: 导出CSV

表 8 正负混合学习性能对比(%，粗体表示最优值)

学习方式		DeepLabV2+ResNet101
正负学习 (不同阈值t)	6	70.96
	7	71.26
	8	71.22

下载: 导出CSV

表 9 六个种子及聚合种子与真值标签在同一像素位置的类别标签一致率和不一致率比较(%，粗体表示最优值)

不同种子	一致率(%)	不一致率(%)
0.05-1.0	89.8000	10.2000
0.05-1.2	89.8000	10.2000
0.10-1.0	89.8711	10.1289
0.10-1.2	89.9034	10.0966
0.20-1.2	89.8732	10.1268
0.30-1.2	89.8284	10.1716
聚合种子	89.9964	10.0036

下载: 导出CSV

表 10 优选标签在不同阈值下标记的干净情况与真值标签的一致率和不一致率比较(%)

不同阈值t	优选的种子标签与真值标签一致
	(a)标记为干净(正确标记)	(b)标记为噪声(错误标记)
3	89.9042	0.0022
4	89.5418	0.3647
5	88.8536	1.0528
6	87.8883	2.0182

下载: 导出CSV

表 11 优选标签在不同阈值下标记的噪声情况与真值标签的一致率和不一致率比较(%)

不同阈值t	优选的种子标签与真值标签不一致
	(c)标记为噪声(正确标记)	(d)标记为干净(错误标记)
3	0.0033	10.0002
4	0.4725	9.6211
5	1.1762	8.9174
6	2.0201	8.0734

下载: 导出CSV

参考文献(46)

[1]	LIU Huan, LI Wei, XIA Xianggen, et al. SegHSI: Semantic segmentation of hyperspectral images with limited labeled pixels[J]. IEEE Transactions on Image Processing, 2024, 33: 6469–6482. doi: 10.1109/TIP.2024.3492724.
[2]	张印辉, 张金凯, 何自芬, 等. 全局感知与稀疏特征关联图像级弱监督病理图像分割[J]. 电子与信息学报, 2024, 46(9): 3672–3682. doi: 10.11999/JEIT240364. ZHANG Yinhui, ZHANG Jinkai, HE Zifen, et al. Global perception and sparse feature associate image-level weakly supervised pathological image segmentation[J]. Journal of Electronics & Information Technology, 2024, 46(9): 3672–3682. doi: 10.11999/JEIT240364.
[3]	LI Jiale, DAI Hang, HAN Hao, et al. MSeg3D: Multi-modal 3D semantic segmentation for autonomous driving[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, 2023: 21694–21704. doi: 10.1109/CVPR52729.2023.02078.
[4]	梁燕, 易春霞, 王光宇, 等. 基于多尺度语义编解码网络的遥感图像语义分割[J]. 电子学报, 2023, 51(11): 3199–3214. doi: 10.12263/DZXB.20220503. LIANG Yan, YI Chunxia, WANG Guangyu, et al. Semantic segmentation of remote sensing image based on multi-scale semantic encoder-decoder network[J]. Acta Electronica Sinica, 2023, 51(11): 3199–3214. doi: 10.12263/DZXB.20220503.
[5]	OH Y, KIM B, and HAM B. Background-aware pooling and noise-aware loss for weakly-supervised semantic segmentation[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nashville, USA, 2021: 6909–6918. doi: 10.1109/CVPR46437.2021.00684.
[6]	LIANG Zhiyuan, WANG Tiancai, ZHANG Xiangyu, et al. Tree energy loss: Towards sparsely annotated semantic segmentation[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, 2022: 16886–16895. doi: 10.1109/CVPR52688.2022.01640.
[7]	ZHAO Yuanhao, SUN Genyun, LING Ziyan, et al. Point-based weakly supervised deep learning for semantic segmentation of remote sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 5638416. doi: 10.1109/TGRS.2024.3409903.
[8]	KWEON H, YOON S H, KIM H, et al. Unlocking the potential of ordinary classifier: Class-specific adversarial erasing framework for weakly supervised semantic segmentation[C]. IEEE/CVF International Conference on Computer Vision, Montreal, Canada, 2021: 6974–6983. doi: 10.1109/ICCV48922.2021.00691.
[9]	ZHOU Bolei, KHOSLA A, LAPEDRIZA A, et al. Learning deep features for discriminative localization[C]. IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 2921–2929. doi: 10.1109/CVPR.2016.319.
[10]	WANG Xiang, YOU Shaodi, LI Xi, et al. Weakly-supervised semantic segmentation by iteratively mining common object features[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, USA, 2018: 1354–1362. doi: 10.1109/CVPR.2018.00147.
[11]	WANG Xun, ZHANG Haozhi, HUANG Weilin, et al. Cross-batch memory for embedding learning[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, USA, 2020: 6387–6396. doi: 10.1109/CVPR42600.2020.00642.
[12]	LEE S, LEE M, LEE J, et al. Railroad is not a train: Saliency as pseudo-pixel supervision for weakly supervised semantic segmentation[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nashville, USA, 2021: 5491–5501. doi: 10.1109/CVPR46437.2021.00545.
[13]	LEE J, OH S J, YUN S, et al. Weakly supervised semantic segmentation using out-of-distribution data[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, 2022: 16876–16885. doi: 10.1109/CVPR52688.2022.01639.
[14]	CHANG Yuting, WANG Qiaosong, HUNG W C, et al. Weakly-supervised semantic segmentation via sub-category exploration[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, USA, 2020: 8988–8997. doi: 10.1109/CVPR42600.2020.00901.
[15]	ARPIT D, JASTRZĘBSKI S, BALLAS N, et al. A closer look at memorization in deep networks[C]. 34th International Conference on Machine Learning, Sydney, Australia, 2017: 233–242.
[16]	CHEN Tao, YAO Yazhou, and TANG Jinhui. Multi-granularity denoising and bidirectional alignment for weakly supervised semantic segmentation[J]. IEEE Transactions on Image Processing, 2023, 32: 2960–2971. doi: 10.1109/TIP.2023.3275913.
[17]	RONG Shenghai, TU Bohai, WANG Zilei, et al. Boundary-enhanced co-training for weakly supervised semantic segmentation[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, 2023: 19574–19584. doi: 10.1109/CVPR52729.2023.01875.
[18]	WU Zifeng, SHEN Chunhua, and VAN DEN HENGEL A. Wider or deeper: Revisiting the ResNet model for visual recognition[J]. Pattern Recognition, 2019, 90: 119–133. doi: 10.1016/j.patcog.2019.01.006.
[19]	SRIVASTAVA N, HINTON G, KRIZHEVSKY A, et al. Dropout: A simple way to prevent neural networks from overfitting[J]. The Journal of Machine Learning Research, 2014, 15(1): 1929–1958.
[20]	LI Mingjia, XIE Binhui, LI Shuang, et al. VBLC: Visibility boosting and logit-constraint learning for domain adaptive semantic segmentation under adverse conditions[C]. 37th AAAI Conference on Artificial Intelligence, Washington, USA, 2023: 8605–8613. doi: 10.1609/aaai.v37i7.26036.
[21]	YANG Guoqing, ZHU Chuang, and ZHANG Yu. A self-training framework based on multi-scale attention fusion for weakly supervised semantic segmentation[C]. IEEE International Conference on Multimedia and Expo, Brisbane, Australia, 2023: 876–881. doi: 10.1109/ICME55011.2023.00155.
[22]	KRÄHENBÜHL P and KOLTUN V. Efficient inference in fully connected CRFs with gaussian edge potentials[C]. 25th International Conference on Neural Information Processing Systems, Granada, Spain, 2011: 109–117.
[23]	LEE J, KIM E, LEE S, et al. FickleNet: Weakly and semi-supervised semantic image segmentation using stochastic inference[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, USA, 2019: 5262–5271. doi: 10.1109/CVPR.2019.00541.
[24]	KIM Y, YIM J, YUN J, et al. NLNL: Negative learning for noisy labels[C]. IEEE/CVF International Conference on Computer Vision, Seoul, Korea (South), 2019: 101–110. doi: 10.1109/ICCV.2019.00019.
[25]	KIM Y, YUN J, SHON H, et al. Joint negative and positive learning for noisy labels[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nashville, USA, 2021: 9437–9446. doi: 10.1109/CVPR46437.2021.00932.
[26]	EVERINGHAM M, VAN GOOL L, WILLIAMS C K I, et al. The PASCAL visual object classes (VOC) challenge[J]. International Journal of Computer Vision, 2010, 88(2): 303–338. doi: 10.1007/s11263-009-0275-4.
[27]	LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft COCO: Common objects in context[C]. 13th European Conference on Computer Vision, Zurich, Switzerland, 2014: 740–755. doi: 10.1007/978-3-319-10602-1_48.
[28]	HARIHARAN B, ARBELÁEZ P, BOURDEV L, et al. Semantic contours from inverse detectors[C]. International Conference on Computer Vision, Barcelon, Spain, 2011: 991–998. doi: 10.1109/ICCV.2011.6126343.
[29]	SHELHAMER E, LONG J, and DARRELL T. Fully convolutional networks for semantic segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(4): 640–651. doi: 10.1109/TPAMI.2016.2572683.
[30]	DENG Jia, DONG Wei, SOCHER R, et al. ImageNet: A large-scale hierarchical image database[C]. IEEE Conference on Computer Vision and Pattern Recognition, Miami, USA, 2009: 248–255. doi: 10.1109/CVPR.2009.5206848.
[31]	CHEN L C, PAPANDREOU G, KOKKINOS I, et al. Semantic image segmentation with deep convolutional nets and fully connected CRFs[C]. 3rd International Conference on Learning Representations, San Diego, USA, 2015: 24–37.
[32]	CHEN L C, PAPANDREOU G, KOKKINOS I, et al. DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834–848. doi: 10.1109/TPAMI.2017.2699184.
[33]	CHEN Zhaozheng, WANG Tan, WU Xiongwei, et al. Class re-activation maps for weakly-supervised semantic segmentation[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, 2022: 959–968. doi: 10.1109/CVPR52688.2022.00104.
[34]	ZHOU Tianfei, ZHANG Meijie, ZHAO Fang, et al. Regional semantic contrast and aggregation for weakly supervised semantic segmentation[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, 2022: 4289–4299. doi: 10.1109/CVPR52688.2022.00426.
[35]	XIE Enze, WANG Wenhai, YU Zhiding, et al. SegFormer: Simple and efficient design for semantic segmentation with transformers[C]. 35th International Conference on Neural Information Processing Systems, 2021: 924. (查阅网上资料, 未找到本条文献出版地信息, 请确认并补充).
[36]	LIU Ze, LIN Yutong, CAO Yue, et al. Swin transformer: Hierarchical vision transformer using shifted windows[C]. IEEE/CVF International Conference on Computer Vision, Montreal, Canada, 2021: 9992–10002. doi: 10.1109/ICCV48922.2021.00986.
[37]	CHENG Bowen, MISRA I, SCHWING A G, et al. Masked-attention mask transformer for universal image segmentation[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, 2022: 1280–1289. doi: 10.1109/CVPR52688.2022.00135.
[38]	HAN W, KANG S, CHOO K, et al. Complementary branch fusing class and semantic knowledge for robust weakly supervised semantic segmentation[J]. Pattern Recognition, 2025, 157: 110922. doi: 10.1016/j.patcog.2024.110922.
[39]	CHEN Qi, YANG Lingxiao, LAI Jianhuang, et al. Self-supervised image-specific prototype exploration for weakly supervised semantic segmentation[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, 2022: 4278–4288. doi: 10.1109/CVPR52688.2022.00425.
[40]	CHEN Liyi, LEI Chenyang, LI Ruihuang, et al. FPR: False positive rectification for weakly supervised semantic segmentation[C]. IEEE/CVF International Conference on Computer Vision, Paris, France, 2023: 1108–1118. doi: 10.1109/ICCV51070.2023.00108.
[41]	CHEN L C, ZHU Yukun, PAPANDREOU G, et al. Encoder-decoder with atrous separable convolution for semantic image segmentation[C]. 15th European Conference on Computer Vision, Munich, Germany, 2018: 833–851. doi: 10.1007/978-3-030-01234-2_49.
[42]	LIU Sheng, LIU Kangning, ZHU Weicheng, et al. Adaptive early-learning correction for segmentation from noisy annotations[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, 2022: 2596–2606. doi: 10.1109/CVPR52688.2022.00263.
[43]	LEE M, LEE S, LEE J, et al. Saliency as pseudo-pixel supervision for weakly and semi-supervised semantic segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(10): 12341–12357. doi: 10.1109/TPAMI.2023.3273592.
[44]	LI Yi, DUAN Yiqun, KUANG Zhanghui, et al. Uncertainty estimation via response scaling for pseudo-mask noise mitigation in weakly-supervised semantic segmentation[C]. 36th AAAI Conference on Artificial Intelligence, Palo Alto, 2022: 1447–1455. doi: 10.1609/aaai.v36i2.20034. (查阅网上资料,未找到本条文献出版地信息,请确认).
[45]	WU Yuanchen, LI Xiaoqiang, LI Jide, et al. DINO is also a semantic guider: Exploiting class-aware affinity for weakly supervised semantic segmentation[C]. 32nd ACM International Conference on Multimedia, Melbourne, Australia, 2024: 1389–1397. doi: 10.1145/3664647.3681710.
[46]	XU Rongtao, WANG Changwei, SUN Jiaxi, et al. Self correspondence distillation for end-to-end weakly-supervised semantic segmentation[C]. 37th AAAI Conference on Artificial Intelligence, Washington, USA, 2023: 3045–3053. doi: 10.1609/aaai.v37i3.25408.