由于数据收集条件以及标注标准的不一致,各个人脸表情数据集之间存在较为明显的领域偏移,从而导致模型在跨域场景下的性能大幅下降。其中,数据收集条件不一致具体表现为收集环境不一致(实验室受控环境 vs 自然非受控环境)和目标人群不一致;标注标准不一致具体表现为不同数据集的标注人员对于表情的理解具有主观性,易受所处地区文化影响。如图 2 所示,我们可以直观地感受到不同人脸表情数据集之间所存在的明显差异。
T. Chen, T. Pu, H. Wu, Y. Xie, L. Liu, L. Lin, "Cross-Domain Facial Expression Recognition: A Unified Evaluation Benchmark and Adversarial Graph Learning", in TPAMI 2021. [PDF]
Y. Xie, T. Chen, T. Pu, H. Wu, L. Lin, "Adversarial Graph Representation Adaptation for Cross-Domain Facial Expression Recognition", in ACM MM 2020. [PDF]
参考文献
[1] P. Lucey, J. F. Cohn, T. Kanade, J. Saragih, Z. Ambadar, and I. Matthews, "The extended cohn-kanade dataset (ck+): A complete dataset for action unit and emotion-specified expression", in CVPR Workshops 2010.
[2] M. Lyons, S. Akamatsu, M. Kamachi, and J. Gyoba, "Coding facial expressions with gabor wavelets", in FG 1998.
[3] A. Dhall, R. Goecke, S. Lucey, and T. Gedeon, "Static facial expression analysis in tough conditions: Data, evaluation protocol and benchmark", in ICCV Workshop 2011.[4] I. J. Goodfellow, D. Erhan, P. L. Carrier, A. Courville, M. Mirza, B. Hamner, W. Cukierski, Y. Tang, D. Thaler, D.-H. Lee et al., "Challenges in representation learning: A report on three machine learning contests,” Neural Networks 2015.
[5] Z. Zhang, P. Luo, C. C. Loy, and X. Tang, “From facial expression recognition to interpersonal relation prediction", in IJCV 2018.
[6] S. Li and W. Deng, “Reliable crowdsourcing and deep locality preserving learning for unconstrained facial expression recognition", in TIP 2018.
[7] Y. Ji, Y. Hu, Y. Yang, F. Shen, and H. T. Shen, "Cross-domain facial expression recognition via an intra-category common feature and inter-category distinction feature fusion network", in Neurocomputing 2019.
[8] R. Zhu, G. Sang, and Q. Zhao, "Discriminative feature adaptation for cross-domain facial expression recognition", in ICB 2016.
[9] S. Li, W. Deng, and J. Du, "Reliable crowdsourcing and deep locality-preserving learning for expression recognition in the wild", CVPR 2017.
[10] S. Li and W. Deng, "Deep emotion transfer network for cross-database facial expression recognition", in ICPR 2018.
[11] M. V. Zavarez, R. F. Berriel, and T. Oliveira-Santos, "Cross-database facial expression recognition based on fine-tuned deep convolutional network", in SIBGRAPI 2017.
[12] S. Li and W. Deng, "A deeper look at facial expression dataset bias", in TAC 2020.
[13] M. Long, Z. Cao, J. Wang, and M. I. Jordan, "Conditional adversarial domain adaptation", in NIPS 2018.
[14] C.-Y. Lee, T. Batra, M. H. Baig, and D. Ulbricht, "Sliced wasserstein discrepancy for unsupervised domain adaptation", in CVPR 2019.
[15] K. Fatras, T. Sejourne, R. Flamary, and N. Courty, "Unbalanced minibatch optimal transport; applications to domain adaptation", in PMLR 2021.
[16] M. Li, Y.-M. Zhai, Y.-W. Luo, P.-F. Ge, and C.-X. Ren, "Enhanced transport distance for unsupervised domain adaptation", in CVPR 2020.
参考文献
[1] P. Lucey, J. F. Cohn, T. Kanade, J. Saragih, Z. Ambadar, and I. Matthews, "The extended cohn-kanade dataset (ck+): A complete dataset for action unit and emotion-specified expression", in CVPR Workshops 2010.
[2] M. Lyons, S. Akamatsu, M. Kamachi, and J. Gyoba, "Coding facial expressions with gabor wavelets", in FG 1998.
[3] A. Dhall, R. Goecke, S. Lucey, and T. Gedeon, "Static facial expression analysis in tough conditions: Data, evaluation protocol and benchmark", in ICCV Workshop 2011.[4] I. J. Goodfellow, D. Erhan, P. L. Carrier, A. Courville, M. Mirza, B. Hamner, W. Cukierski, Y. Tang, D. Thaler, D.-H. Lee et al., "Challenges in representation learning: A report on three machine learning contests,” Neural Networks 2015.
[5] Z. Zhang, P. Luo, C. C. Loy, and X. Tang, “From facial expression recognition to interpersonal relation prediction", in IJCV 2018.
[6] S. Li and W. Deng, “Reliable crowdsourcing and deep locality preserving learning for unconstrained facial expression recognition", in TIP 2018.
[7] Y. Ji, Y. Hu, Y. Yang, F. Shen, and H. T. Shen, "Cross-domain facial expression recognition via an intra-category common feature and inter-category distinction feature fusion network", in Neurocomputing 2019.
[8] R. Zhu, G. Sang, and Q. Zhao, "Discriminative feature adaptation for cross-domain facial expression recognition", in ICB 2016.
[9] S. Li, W. Deng, and J. Du, "Reliable crowdsourcing and deep locality-preserving learning for expression recognition in the wild", CVPR 2017.
[10] S. Li and W. Deng, "Deep emotion transfer network for cross-database facial expression recognition", in ICPR 2018.
[11] M. V. Zavarez, R. F. Berriel, and T. Oliveira-Santos, "Cross-database facial expression recognition based on fine-tuned deep convolutional network", in SIBGRAPI 2017.
[12] S. Li and W. Deng, "A deeper look at facial expression dataset bias", in TAC 2020.
[13] M. Long, Z. Cao, J. Wang, and M. I. Jordan, "Conditional adversarial domain adaptation", in NIPS 2018.
[14] C.-Y. Lee, T. Batra, M. H. Baig, and D. Ulbricht, "Sliced wasserstein discrepancy for unsupervised domain adaptation", in CVPR 2019.
[15] K. Fatras, T. Sejourne, R. Flamary, and N. Courty, "Unbalanced minibatch optimal transport; applications to domain adaptation", in PMLR 2021.
[16] M. Li, Y.-M. Zhai, Y.-W. Luo, P.-F. Ge, and C.-X. Ren, "Enhanced transport distance for unsupervised domain adaptation", in CVPR 2020.