AgeFlow: Conditional Age Progression and Regression with Normalizing Flows

Age progression and regression aim to synthesize photorealistic appearance of a given face image with aging and rejuvenation effects, respectively. Existing generative adversarial networks (GANs) based methods suffer from the following three major issues: 1) unstable training introducing strong ghost artifacts in the generated faces, 2) unpaired training leading to unexpected changes in facial attributes such as genders and races, and 3) non-bijective age mappings increasing the uncertainty in the face transformation. To overcome these issues, this paper proposes a novel framework, termed AgeFlow, to integrate the advantages of both flow-based models and GANs. The proposed AgeFlow contains three parts: an encoder that maps a given face to a latent space through an invertible neural network, a novel invertible conditional translation module (ICTM) that translates the source latent vector to target one, and a decoder that reconstructs the generated face from the target latent vector using the same encoder network; all parts are invertible achieving bijective age mappings. The novelties of ICTM are two-fold. First, we propose an attribute-aware knowledge distillation to learn the manipulation direction of age progression while keeping other unrelated attributes unchanged, alleviating unexpected changes in facial attributes. Second, we propose to use GANs in the latent space to ensure the learned latent vector indistinguishable from the real ones, which is much easier than traditional use of GANs in the image domain. Experimental results demonstrate superior performance over existing GANs-based methods on two benchmarked datasets. The source code is available at https://github.com/Hzzone/AgeFlow.

[1]  Yuichi Yoshida,et al.  Spectral Normalization for Generative Adversarial Networks , 2018, ICLR.

[2]  Yang Song,et al.  Age Progression/Regression by Conditional Adversarial Autoencoder , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  Anil K. Jain,et al.  Learning Face Age Progression: A Pyramid Architecture of GANs , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[4]  Stefanos Zafeiriou,et al.  ArcFace: Additive Angular Margin Loss for Deep Face Recognition , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[5]  Samy Bengio,et al.  Density estimation using Real NVP , 2016, ICLR.

[6]  Yu Qiao,et al.  Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks , 2016, IEEE Signal Processing Letters.

[7]  Xiaogang Wang,et al.  Deep Learning Face Attributes in the Wild , 2014, 2015 IEEE International Conference on Computer Vision (ICCV).

[8]  Roy Or-El,et al.  Lifespan Age Transformation Synthesis , 2020, ECCV.

[9]  Simon Osindero,et al.  Conditional Generative Adversarial Nets , 2014, ArXiv.

[10]  Prafulla Dhariwal,et al.  Glow: Generative Flow with Invertible 1x1 Convolutions , 2018, NeurIPS.

[11]  Karl Ricanek,et al.  MORPH: a longitudinal image database of normal adult age-progression , 2006, 7th International Conference on Automatic Face and Gesture Recognition (FGR06).

[12]  Shiguang Shan,et al.  S2GAN: Share Aging Factors Across Ages and Share Aging Trends Among Individuals , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[13]  Yiying Tong,et al.  Age-Invariant Face Recognition , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[14]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[15]  Yoshua Bengio,et al.  NICE: Non-linear Independent Components Estimation , 2014, ICLR.

[16]  Hongming Shan,et al.  When Age-Invariant Face Recognition Meets Face Age Synthesis: A Multi-Task Learning Framework , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[17]  Zhenan Sun,et al.  Global and Local Consistent Wavelet-Domain Age Synthesis , 2018, IEEE Transactions on Information Forensics and Security.

[18]  Heng Tao Shen,et al.  Dual Conditional GANs for Face Aging and Rejuvenation , 2018, IJCAI.

[19]  Zhenan Sun,et al.  Attribute-Aware Face Aging With Wavelet-Based Generative Adversarial Networks , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[20]  Zhenan Sun,et al.  Hierarchical Face Aging Through Disentangled Latent Characteristics , 2020, ECCV.

[21]  Raymond Y. K. Lau,et al.  Least Squares Generative Adversarial Networks , 2016, 2017 IEEE International Conference on Computer Vision (ICCV).

[22]  Xu Tang,et al.  Face Aging with Identity-Preserved Conditional Generative Adversarial Networks , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[23]  Hongming Shan,et al.  PFA-GAN: Progressive Face Aging With Generative Adversarial Network , 2020, IEEE Transactions on Information Forensics and Security.

[24]  Chu-Song Chen,et al.  Face Recognition and Retrieval Using Cross-Age Reference Coding With Cross-Age Celebrity Dataset , 2015, IEEE Transactions on Multimedia.

[25]  Nicu Sebe,et al.  Recurrent Face Aging , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[26]  Qi Li,et al.  Global and Local Consistent Age Generative Adversarial Networks , 2018, 2018 24th International Conference on Pattern Recognition (ICPR).

[27]  Hongming Shan,et al.  Routinggan: Routing Age Progression and Regression with Disentangled Learning , 2021, ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[28]  拓海 杉山,et al.  “Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告 , 2017 .

[29]  Qi Li,et al.  Age Progression and Regression with Spatial Attention Modules , 2019, AAAI.