Deep Identity-aware Transfer of Facial Attributes

By Mu Li, Wangmeng Zuo, David Zhang. 2016

153 citations recorded by the source corpus.

This paper presents a Deep convolutional network model for Identity-Aware Transfer (DIAT) of facial attributes. Given the source input image and the reference attribute, DIAT aims to generate a facial image that owns the reference attribute as well as keeps the same or similar identity to the input image. In general, our model consists of a mask network and an attribute transform network which work in synergy to generate a photo-realistic facial image with the reference attribute. Considering that the reference attribute may be only related to some parts of the image, the mask network is introduced to avoid the incorrect editing on attribute irrelevant region. Then the estimated mask is adopted to combine the input and transformed image for producing the transfer result. For joint training of transform network and mask network, we incorporate the adversarial attribute loss, identity-aware adaptive perceptual loss, and VGG-FACE based identity loss. Furthermore, a denoising network is presented to serve for perceptual regularization to suppress the artifacts in transfer result, while an attribute ratio regularization is introduced to constrain the size of attribute relevant region. Our DIAT can provide a unified solution for several representative facial attribute transfer tasks, e.g., expression transfer, accessory removal, age progression, and gender transfer, and can be extended for other face enhancement tasks such as face hallucination. The experimental results validate the effectiveness of the proposed method. Even for the identity-related attribute (e.g., gender), our DIAT can obtain visually impressive results by changing the attribute while retaining most identity-aware features.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

StarGAN: Unified Generative Adversarial Networks for Multi-domain Image-to-Image Translation (2017, methodology): “As our baseline models, we adopt DIAT [15] and CycleGAN [32], both of which performs image-to-image translation between two different domains.”
AttGAN: Facial Attribute Editing by Only Changing What You Want (2017, methodology): “[9] present to train a deep identity-aware attribute transfer model to add/remove an attribute to/from a face image by employing”
Generative adversarial interpolative autoencoding: adversarial training on latent space interpolations encourage convex latent distributions (2018, background): “…(e.g. Radford et al. 2015; Larsen et al. 2015; Kingma & Dhariwal 2018; White 2016), and (2) supervised and partially supervised approaches, where high-level feature information is used during learning (e.g. Choi et al. 2017; Isola et al. 2017; Li et al. 2016; Perarnau et al. 2016; Zhu et al. 2017).”
GANimation: Anatomically-aware Facial Animation from a Single Image (2018, methodology): “Facial Expression Synthesis results for: DIAT [20], CycleGAN [28], IcGAN [26] and StarGAN [4]; and ours.”
Learning Residual Images for Face Attribute Manipulation (2016, methodology): “(1) Our method generates manipulated images using residual images which is different from [19].”
DeepFake Detection by Analyzing Convolutional Traces (2020, result): “Results have been compared with other existing methods [26, 30, 45] and showed how STARGAN manages to generate images of superior visual quality.”
Model Compression with Multi-Task Knowledge Distillation for Web-scale Question Answering System (2019, methodology): “Transfer learning is a method to transfer knowledge from one task to another, which has been widely used in various ﬁelds (Li et al., 2016; Murez et al., 2018; Bansal et al., 2018; He et al., 2016).”
A survey on face data augmentation for the training of deep neural networks (2019, methodology): “The typical examples of autoregressive generative models are PixelRNN and PixelCNN proposed by Van Den Oord et al. [137].”
Arbitrary Facial Attribute Editing: Only Change What You Want (2017, methodology): “[12] present a deep identity-aware attribute transfer (DIAT) model to add/remove an attribute to/from a face image via adversarial learning.”
MaskGAN: Towards Diverse and Interactive Facial Image Manipulation (2019, methodology): “DIAT [23] utilized adversarial loss to transfer attributes and learn to blend predicted face and original face.”
Every Smile is Unique: Landmark-Guided Diverse Smile Generation (2018, background): “Similarly, in [24] a GAN-based model is proposed for transferring facial attributes while preserving as much as possible information about identity.”
STGAN: A Unified Selective Transfer Network for Arbitrary Image Attribute Editing (2019, methodology): “While several methods have been proposed to learn single translation model for each specific attribute editing task [5, 20, 29, 32], they suffer from the limitation of image-to-image translation and cannot well scale to arbitrary attribute editing.”