《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (8): 2381-2386.DOI: 10.11772/j.issn.1001-9081.2023081173
收稿日期:
2023-09-01
修回日期:
2023-10-19
接受日期:
2023-11-03
发布日期:
2024-08-22
出版日期:
2024-08-10
通讯作者:
胡春龙
作者简介:
孙淳(1999—),男,江苏宿迁人,硕士研究生,CCF会员,主要研究方向:计算机视觉、机器学习基金资助:
Chun SUN, Chunlong HU(), Shucheng HUANG
Received:
2023-09-01
Revised:
2023-10-19
Accepted:
2023-11-03
Online:
2024-08-22
Published:
2024-08-10
Contact:
Chunlong HU
About author:
SUN Chun , born in 1999, M. S. candidate. His research interestsinclude computer vision, machine learning.Supported by:
摘要:
在基于传统的排序、回归的年龄估计方法中,存在不能有效利用人脸的演化特征、构建不同排序标签之间的相关性,且二分类方法进行年龄估计会产生排序不一致的问题。基于上述问题,提出一致性保留的集成排序年龄估计方法,充分利用年龄与排序值之间的相关性,抑制排序不一致问题;并提出新指标——排序不一致样本比例,用于评估二分类排序方法中排序不一致问题。首先,通过设计的编码方法将年龄类别转换成排序矩阵形式;然后,使用残差网络ResNet34(Residual Network)特征提取网络提取面部特征,再通过提出的编码学习模块进行编码学习;最后,通过基于度量方法的排序解码器将网络预测结果解码成图片的预测年龄。在MORPH Ⅱ数据集上的实验结果达到平均绝对误差(MAE)2.18,并在其他公开数据集上与同样基于排序、有序回归方法的OR-CNN(Ordinal Regression with CNN)、CORAL(COnsistent RAnk Logits)等方法相比,所提方法有更准确的预测结果,同时抑制了排序不一致样本的比例,排序不一致度量性能比OR-CNN方法提升了约65%。
中图分类号:
孙淳, 胡春龙, 黄树成. 一致性保留的集成排序年龄估计方法[J]. 计算机应用, 2024, 44(8): 2381-2386.
Chun SUN, Chunlong HU, Shucheng HUANG. Consistency preserving age estimation method by ensemble ranking[J]. Journal of Computer Applications, 2024, 44(8): 2381-2386.
随机种子数 | MAE | ||
---|---|---|---|
1 | 2.17 | 80.0 | 92.7 |
2 | 2.18 | 79.5 | 92.8 |
3 | 2.18 | 79.7 | 92.6 |
4 | 2.18 | 79.4 | 92.5 |
5 | 2.18 | 79.2 | 92.5 |
表1 80-20随机划分方式下MORPH Ⅱ数据集MAE、CS(3)、CS(5)结果
Tab. 1 MAE, CS(3), CS(5) results of MORPH Ⅱ dataset by 80-20 random partition method
随机种子数 | MAE | ||
---|---|---|---|
1 | 2.17 | 80.0 | 92.7 |
2 | 2.18 | 79.5 | 92.8 |
3 | 2.18 | 79.7 | 92.6 |
4 | 2.18 | 79.4 | 92.5 |
5 | 2.18 | 79.2 | 92.5 |
方法 | MORPH Ⅱ | CACD | AFAD | FG-NET |
---|---|---|---|---|
OR-CNN[ | 3.27 | 5.52 | 3.68 | — |
SSR-Net [ | 3.16 | — | — | — |
Ranking-CNN[ | 2.96 | — | — | — |
DOEL[ | 2.81 | — | — | 3.44 |
Group-n[ | 2.52 | 4.68 | — | 2.96 |
DEX[ | 2.68 | 4.79 | — | 4.30 |
ALD-Net[ | 2.65 | 4.62 | — | 3.25 |
CORAL[ | 2.64 | 5.39 | 3.49 | — |
FCRN[ | 2.72 | — | 3.28 | — |
DCDL+MV[ | 2.45 | — | 2.28 | — |
MV[ | — | — | 4.10 | |
本文方法 | 2.18 | 3.04 |
表2 在不同数据集上使用RS数据集划分方法得到的MAE结果
Tab. 2 MAE results obtained by RS dataset partitioning method on different datasets
方法 | MORPH Ⅱ | CACD | AFAD | FG-NET |
---|---|---|---|---|
OR-CNN[ | 3.27 | 5.52 | 3.68 | — |
SSR-Net [ | 3.16 | — | — | — |
Ranking-CNN[ | 2.96 | — | — | — |
DOEL[ | 2.81 | — | — | 3.44 |
Group-n[ | 2.52 | 4.68 | — | 2.96 |
DEX[ | 2.68 | 4.79 | — | 4.30 |
ALD-Net[ | 2.65 | 4.62 | — | 3.25 |
CORAL[ | 2.64 | 5.39 | 3.49 | — |
FCRN[ | 2.72 | — | 3.28 | — |
DCDL+MV[ | 2.45 | — | 2.28 | — |
MV[ | — | — | 4.10 | |
本文方法 | 2.18 | 3.04 |
方法 | MAE | |
---|---|---|
CORAL[ | 2.64 | 0.000 |
OR-CNN[ | 3.19 | 0.240 |
本文方法 | 2.18 | 0.083 |
表3 本文方法与同类方法在MORPH Ⅱ数据集上排序不一致度量对比
Tab. 3 Comparison of inconsistent ranking measures between proposed method and similar methods on MORPH Ⅱ dataset
方法 | MAE | |
---|---|---|
CORAL[ | 2.64 | 0.000 |
OR-CNN[ | 3.19 | 0.240 |
本文方法 | 2.18 | 0.083 |
1 | 曹猛,田青,马廷淮,等.人类面部属性估计研究:综述[J]. 软件学报, 2019, 30(7):2188-2207. |
CAO M, TIAN Q, MA T H, et al. Human facial attributes estimation: a survey [J]. Journal of Software, 2019, 30(7): 2188-2207. | |
2 | GUNAY A, NABIYEV V V. Automatic age classification with LBP[C]// Proceedings of the 2008 23rd Intnational Symposium on Computer and Information Sciences. Piscataway: IEEE, 2008:1-4. |
3 | GUO G, MU G W, FU Y, et al. Human age estimation using bio-inspired features[C]// Proceedings of the 2009 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2009: 112-119. |
4 | GUO G, FU Y, DYER C R, et al. Image-based human age estimation by manifold learning and locally adjusted robust regression[J]. IEEE Transactions on Image Processing, 2008, 17(7): 1178-1788. |
5 | YANG P, ZHONG L, METAXAS D. Ranking model for facial age estimation[C]// Proceedings of the 2010 20th International Conference on Pattern Recognition. Piscataway: IEEE, 2010: 3404-3407. |
6 | CHANG K-Y, CHEN C-S, Y-P HUNG. Ordinal hyperplanes ranker with cost sensitivities for age estimation[C]// Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2011: 585-592. |
7 | CAO D, LEI Z, ZHANG Z, et al. Human age estimation using ranking SVM[C]// Proceedings of the 7th Chinese Conference on Biometric Recognition. Berlin: Springer, 2012: 324-331. |
8 | LI C, LIU Q, LIU J, et al. Learning ordinal discriminative features for age estimation[C]// Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2012: 2570-2577. |
9 | 何建辉,胡春龙,束鑫.基于多峰标签分布学习的多任务年龄估计方法[J].计算机应用, 2023, 43(5): 1578-1583. |
HE J H, HU C L, SHU X. Multi-task age estimation method based on multi-peak label distribution learning [J]. Journal of Computer Applications, 2023, 43(5): 1578-1583. | |
10 | 李睿钰,祝继华,刘新媛. 考虑标记间协作的标记分布学习[J]. 软件学报,2022,33(2):539-554. |
LI R Y, ZHU J H, LIU X Y. Label distribution learning with collaboration among labels[J]. Journal of Software, 2022, 33(2): 539-554. | |
11 | HU C, GAO J, CHEN J, et al. Fine-grained age estimation with multi-attention network[J]. IEEE Access, 2020, 8: 196013-196023. |
12 | ROTHE R, TIMOFTE R, VAN GOOL L. DEX: deep expectation of apparent age from a single image[C]// Proceedings of the 2015 IEEE International Conference on Computer Vision Workshop. Piscataway: IEEE, 2015: 252-257. |
13 | LIU X, LI S, KAN M, et al. AgeNet: deeply learned regressor and classifier for robust apparent age estimation[C]// Proceedings of the 2015 IEEE International Conference on Computer Vision Workshop. Piscataway: IEEE, 2015: 258-266. |
14 | MALLI R C, AYGÜN M, EKENEL H K. Apparent age estimation using ensemble of deep learning models[C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition Workshops. Piscataway: IEEE, 2016: 714-721. |
15 | YANG T-Y, HUANG Y-H, LIN Y-Y, et al. SSR-Net: a compact soft stagewise regression network for age estimation[C]// Proceedings of the 27th International Joint Conference on Artificial Intelligence. California: ijcai.org, 2018: 1078-1084. |
16 | XIE J-C, C-M PUN. Chronological age estimation under the guidance of age-related facial attributes [J]. IEEE Transactions on Information Forensics and Security, 2019, 14(9): 2500-2511. |
17 | PAN H, HAN H, SHAN S, et al. Mean-variance loss for deep age estimation from a face[C]// Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 5285-5294. |
18 | LIU N, SU Z, CHAI Y, et al. Feedback neural network for constrained biobjective convex optimization [J]. Neurocomputing, 2022, 514: 127-136. |
19 | LIU H, LU J, FENG J,et al. Ordinal deep feature learning for facial age estimation[C]// Proceedings of the 2017 12th IEEE International Conference on Automatic Face & Gesture Recognition. Piscataway: IEEE, 2017: 157-164. |
20 | LI K, XING J, HU W, et al. D2C: deep cumulatively and comparatively learning for human age estimation[J]. Pattern Recognition, 2017, 66: 95-105. |
21 | SUN H, PAN H, HAN H, et al. Deep conditional distribution learning for age estimation[J]. IEEE Transactions on Information Forensics and Security, 2021, 16: 4679-4690. |
22 | HE J, HU C, WANG L. Facial age estimation based on asymmetrical label distribution[J]. Multimedia Systems,2023,29: 753-762. |
23 | GAO B-B, XING C, XIE C-W, et al. Deep label distribution learning with label ambiguity[J]. IEEE Transactions on Image Processing, 2017, 26(6): 2825-2838. |
24 | CHEN S, ZHANG C, DONG M, et al. Using ranking-CNN for age estimation[C]// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017: 742-751. |
25 | NIU Z, ZHOU M, WANG L, et al. Ordinal regression with multiple output CNN for age estimation[C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2016: 4920-4928. |
26 | CAO W, MIRJALILI V, RASCHKA S. Rank consistent ordinal regression for neural networks with application to age estimation[J]. Pattern Recognition Letters, 2020, 140: 325-331. |
27 | RICANEK K, TESAFAYE T. MORPH: a longitudinal image database of normal adult age-progression[C]// Proceedings of the 7th International Conference on Automatic Face and Gesture Recognition. Piscataway: IEEE, 2006: 341-345. |
28 | PANIS G, LANITIS A, TSAPATSOULIS N, et al. Overview of research on facial ageing using the fg-net ageing database [J]. IET Biometrics,2 016, 5(2): 37-46. |
29 | CHEN B-C, CHEN C-S, HSU W H. Face recognition and retrieval using cross-age reference coding with cross-age celebrity dataset [J]. IEEE Transactions on Multimedia, 2015, 17(6): 804-815. |
30 | XIE J-C, C-M PUN. Deep and ordinal ensemble learning for human age estimation from facial images [J]. IEEE Transactions on Information Forensics and Security, 2020, 15: 2361-2374. |
31 | TAN Z, WAN J, LEI Z, et al. Efficient group-n encoding and decoding for facial age estimation [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(11): 2610-2623. |
32 | CHEN G, PENG J, WANG L, et al. Feature constraint reinforcement based age estimation[J]. Multimedia Tools and Applications, 2023, 82: 17033-17054. |
[1] | 孙晓玲, 王丹辉, 李姗姗. 基于区块链的动态密文排序检索方案[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2500-2505. |
[2] | 付帅, 郭小英, 白茹意, 闫涛, 陈斌. 改进的CloFormer模型与有序回归相结合的年龄评估方法[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2372-2380. |
[3] | 孟繁珺, 韩斌, 黄树成, 梅向东. 用于时态聚合范围查询的分布式时态索引[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1848-1854. |
[4] | 陈美宏, 袁凌云, 夏桐. 基于主从多链的数据分类分级访问控制模型[J]. 《计算机应用》唯一官方网站, 2024, 44(4): 1148-1157. |
[5] | 王元龙, 胡文博, 张虎. 知识引导的视觉关系检测模型[J]. 《计算机应用》唯一官方网站, 2024, 44(3): 683-689. |
[6] | 张祖篡, 陈学斌, 高瑞, 邹元怀. 基于标签分类的联邦学习客户端选择方法[J]. 《计算机应用》唯一官方网站, 2024, 44(12): 3759-3765. |
[7] | 郑升旻, 付晓东. 利用混合Plackett-Luce模型的不完整序数偏好预测[J]. 《计算机应用》唯一官方网站, 2024, 44(10): 3105-3113. |
[8] | 马勇健, 史旭华, 王佩瑶. 基于两阶段搜索与动态资源分配的约束多目标进化算法[J]. 《计算机应用》唯一官方网站, 2024, 44(1): 269-277. |
[9] | 彭诗杰, 陈红梅, 王丽珍, 肖清. 基于地理偏好排序的兴趣点混合推荐模型[J]. 《计算机应用》唯一官方网站, 2023, 43(8): 2448-2455. |
[10] | 刘耀, 童昕, 陈一风. 面向业务需求的算法路径自组配模型[J]. 《计算机应用》唯一官方网站, 2023, 43(6): 1768-1778. |
[11] | 何建辉, 胡春龙, 束鑫. 基于多峰标签分布学习的多任务年龄估计方法[J]. 《计算机应用》唯一官方网站, 2023, 43(5): 1578-1583. |
[12] | 王彬, 向甜, 吕艺东, 王晓帆. 基于NSGA‑Ⅱ的自适应多尺度特征通道分组优化算法[J]. 《计算机应用》唯一官方网站, 2023, 43(5): 1401-1408. |
[13] | 张敏, 韩晓龙. 多目标模糊机会约束规划的低碳多式联运路径优化[J]. 《计算机应用》唯一官方网站, 2023, 43(2): 636-644. |
[14] | 薛海蓉, 韩晓龙. 基于改进NSGA-Ⅱ的考虑自动引导车充电策略的集成调度[J]. 《计算机应用》唯一官方网站, 2023, 43(12): 3848-3855. |
[15] | 李二超, 张生辉. 基于新评价指标自适应预测的动态多目标优化算法[J]. 《计算机应用》唯一官方网站, 2023, 43(10): 3178-3187. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||