基于自注意力机制的亲属关系识别
作者:
中图分类号:

TP391.4

基金项目:

四川理工学院科研基金(2019RC12); 四川省人工智能重点实验室开放基金(2020RZJ03)


Kinship Recognition Based on Self-attention Mechanism
Author:
  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献 [20]
  • |
  • 相似文献 [20]
  • | | |
  • 文章评论
    摘要:

    目前, 基于局部注意力机制的卷积神经网络(CNNs)用于亲属关系识别特征提取获得了不错的效果, 但基于卷积神经网络的主干模型提升不明显, 同时鲜有研究者使用具有全局信息捕获能力的自注意机制. 因此, 提出使用基于一种无卷积主干特征提取网络的S-ViT模型, 即用具有自全局注意力机制的Vision Transformer作为基础主干特征提取网络, 通过构建孪生网络与具有局部注意力机制的CNN相结合, 扩大传统分类网络, 用于亲属关系识别相关问题的研究. 最终实验结果表明, 相比RFIW2020挑战赛领先的方法, 所提出的方法在亲属关系识别3个任务上获得了良好的效果, 第1个任务中获得了76.8%验证精度排名第二, 第2个和第3个任务中排名第三, 证明了该方法的可行性和有效性, 为亲属关系识别提出了一种新的解决方法.

    Abstract:

    At present, convolutional neural networks (CNNs) based on local attention mechanism have yielded sound results in feature extraction of kinship recognition. However, the improvement of backbone models based on CNNs is not obvious, and few researchers employ self-attention mechanisms with global information capture ability. Therefore, an S-ViT model based on a convolution-free backbone feature extraction network is proposed, which is to adopt Vision Transformer with a self-global attention mechanism as the basic backbone feature extraction network. By constructing a twin network and a CNN with a local attention mechanism, the traditional classification network is expanded for research on related issues of kinship recognition. The final experimental results show that compared with the leading method of the RFIW2020 Challenge, the proposed method has performed well in the three kinship recognition tasks. The first task ranks second with verification accuracy of 76.8%, and the second and third tasks rank third. As a result, the feasibility and effectiveness of the method are improved to propose a new solution to kinship recognition.

    参考文献
    [1] Dornaika F, Arganda-Carreras I, Serradilla O. Transfer learning and feature fusion for kinship verification. Neural Computing and Applications, 2020, 32(11): 7139–7151. [doi: 10.1007/s00521-019-04201-0
    [2] Li ZW, Liu F, Yang WJ, et al. A survey of convolutional neural networks: Analysis, applications, and prospects. IEEE Transactions on Neural Networks and Learning Systems, 2022, 33(12): 6999–7019. [doi: 10.1109/TNNLS.2021.3084827
    [3] Robinson JP, Shao M, Wu Y, et al. Visual kinship recognition of families in the wild. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(11): 2624–2637. [doi: 10.1109/TPAMI.2018.2826549
    [4] Robinson JP, Qin C, Shao M, et al. The 5th recognizing families in the wild data challenge: Predicting kinship from faces. Proceedings of the 16th IEEE International Conference on Automatic Face and Gesture Recognition. Jodhpur: IEEE, 2021. 1–7.
    [5] Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words: Transformers for image recognition at scale. Proceedings of the 9th International Conference on Learning Representations. OpenReview.net, 2021.
    [6] Chu XX, Tian Z, Wang YQ, et al. Twins: Revisiting the design of spatial attention in vision transformers. Proceedings of the 35th International Conference on Neural Information Processing Systems. 2021. 9355–9366.
    [7] Chu XX, Tian Z, Zhang B, et al. Conditional positional encodings for vision transformers. arXiv:2102.10882, 2021.
    [8] Wang WH, Xie EZ, Li X, et al. Pyramid vision transformer: A versatile backbone for dense prediction without convolutions. Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021. 568–578.
    [9] Bromley J, Guyon I, LeCun Y, et al. Signature verification using a “Siamese” time delay neural network. Proceedings of the 6th International Conference on Neural Information Processing Systems. Denver: Morgan Kaufmann Publishers Inc., 1993. 737–744.
    [10] Deng J, Dong W, Socher R, et al. ImageNet: A large-scale hierarchical image database. Proceedings of the 2009 IEEE Conference on Computer Vision and Pattern Recognition. Miami: IEEE, 2009. 248–255.
    [11] Tan CQ, Sun FC, Kong T, et al. A survey on deep transfer learning. Proceedings of the 27th International Conference on Artificial Neural Networks. Rhodes: Springer, 2018. 270–279.
    [12] de Boer PT, Kroese DP, Mannor S, et al. A tutorial on the cross-entropy method. Annals of Operations Research, 2005, 134(1): 19–67. [doi: 10.1007/s10479-005-5724-z
    [13] Wen YD, Zhang KP, Li ZF, et al. A discriminative feature learning approach for deep face recognition. Proceedings of the 14th European Conference on Computer Vision. Amsterdam: Springer, 2016. 499–515.
    [14] Liu Z, Lin YT, Cao Y, et al. Swin Transformer: Hierarchical vision transformer using shifted windows. Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021. 10012–10022.
    [15] Khosla P, Teterwak P, Wang C, et al. Supervised contrastive learning. Proceedings of the 33rd International Conference on Neural Information Processing Systems. 2020. 18661–18673.
    [16] Hoffer E, Ailon N. Deep metric learning using triplet network. Proceedings of the 3rd International Workshop on Similarity-based Pattern Recognition. Copenhagen: Springer, 2015. 84–92.
    [17] Yu J, Li MY, Hao XL, et al. Deep fusion Siamese network for automatic kinship verification. Proceedings of the 15th IEEE International Conference on Automatic Face and Gesture Recognition. Buenos Aires: IEEE, 2020. 892–899.
    [18] Luo ZP, Zhang ZG, Xu ZY, et al. Challenge report recognizing families in the wild data challenge. Proceedings of the 15th IEEE International Conference on Automatic Face and Gesture Recognition. Buenos Aires: IEEE, 2020. 868–871.
    [19] Dahan E, Keller Y. A unified approach to kinship verification. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(8): 2851–2857
    [20] Huang JY, Strome MB, Jenkins I, et al. Solving the families in the wild kinship verification challenge by program synthesis. Proceedings of the 16th IEEE International Conference on Automatic Face and Gesture Recognition. Jodhpur: IEEE, 2021. 1–5.
    引证文献
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

李德财,蒋行国,何李,李嘉莉.基于自注意力机制的亲属关系识别.计算机系统应用,2023,32(9):89-96

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2022-12-21
  • 最后修改日期:2023-01-09
  • 在线发布日期: 2023-07-14
文章二维码
您是第11209279位访问者
版权所有:中国科学院软件研究所 京ICP备05046678号-3
地址:北京海淀区中关村南四街4号 中科院软件园区 7号楼305房间,邮政编码:100190
电话:010-62661041 传真: Email:csa (a) iscas.ac.cn
技术支持:北京勤云科技发展有限公司

京公网安备 11040202500063号