Name
Abstract | ||
|---|---|---|
Facial attribute editing aims to manipulate single or multiple attributes on a given face image, i.e., to generate a new face image with desired attributes while preserving other details. Recently, the generative adversarial net (GAN) and encoder–decoder architecture are usually incorporated to handle this task with promising results. Based on the encoder–decoder architecture, facial attribute editing is achieved by decoding the latent representation of a given face conditioned on the desired attributes. Some existing methods attempt to establish an attribute-independent latent representation for further attribute editing. However, such attribute-independent constraint on the latent representation is excessive because it restricts the capacity of the latent representation and may result in information loss, leading to over-smooth or distorted generation. Instead of imposing constraints on the latent representation, in this work, we propose to apply an
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">attribute classification constraint</italic>
to the generated image to just guarantee the correct change of desired attributes, i.e., to change what you want. Meanwhile, the
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">reconstruction learning</italic>
is introduced to preserve attribute-excluding details, in other words, to only change what you want. Besides, the
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">adversarial learning</italic>
is employed for visually realistic editing. These three components cooperate with each other forming an effective framework for high quality facial attribute editing, referred as
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AttGAN</italic>
. Furthermore, the proposed method is extended for
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">attribute style manipulation</italic>
in an unsupervised manner. Experiments on two wild datasets, CelebA and LFW, show that the proposed method outperforms the state-of-the-art on realistic attribute editing with other facial details well preserved. |
Year | DOI | Venue |
|---|---|---|
2017 | 10.1109/TIP.2019.2916751 | IEEE Transactions on Image Processing |
Keywords | Field | DocType |
Face,Facial features,Task analysis,Decoding,Image reconstruction,Hair,Gallium nitride | Architecture,Information loss,Pattern recognition,Computer science,Artificial intelligence,Generative grammar,Decoding methods,Distortion,Machine learning,Adversarial system | Journal |
Volume | Issue | ISSN |
28 | 11 | 1057-7149 |
Citations | PageRank | References |
27 | 0.93 | 23 |
Authors | ||
5 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Zhenliang He | 1 | 53 | 2.99 |
| Wangmeng Zuo | 2 | 3833 | 173.11 |
| Meina Kan | 3 | 713 | 26.32 |
| Shiguang Shan | 4 | 6322 | 283.75 |
| Xilin Chen | 5 | 6291 | 306.27 |