A Unified Feature Disentangler for Multi-Domain Image Translation and Manipulation

By Alexander H. Liu, Yen-Cheng Liu, Yu-Ying Yeh, Yu-Chiang Frank Wang. 2018

232 citations recorded by the source corpus.

We present a novel and unified deep learning framework which is capable of learning domain-invariant representation from data across multiple domains. Realized by adversarial training with additional ability to exploit domain-specific information, the proposed network is able to perform continuous cross-domain image translation and manipulation, and produces desirable output images accordingly. In addition, the resulting feature representation exhibits superior performance of unsupervised domain adaptation, which also verifies the effectiveness of the proposed model in learning disentangled features for describing cross-domain data.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Generative Models

How researchers see this paper

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

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Generative Models

Citations in the corpus

GAN Inversion: A Survey (2021, background): “[78] propose a unified model that learns disentangled representation for describing and manipulating data across multiple domains, and Lu et al.”
Recent Advances in Autoencoder-Based Representation Learning (2018, background): “While this task as such does not a priori encourage learning of meaningful representation, adding appropriate pressure does: Sharing parts of the latent representation between the translation networks [27–29] and/or combining domain specific and shared translation networks [88] leads to disentangled”
Deep Sketch-Guided Cartoon Video Inbetweening (2021, background): “[34] propose a model to learn disentangled features for describing cross-domain data to perform continuous crossdomain image translation and manipulation.”
Generalizing to Unseen Domains: A Survey on Domain Generalization (2021, unknown): “Qiao et al. [40] also used VAE for disentanglement, where they proposed a Unified Feature Disentanglement Network (UFDN) that treated both data domains and image attributes of interest as latent factors to be disentangled.”
Domain2Vec: Domain Embedding for Unsupervised Domain Adaptation (2020, methodology): “Various methods have been proposed, including discrepancybased UDA approaches [2,30,31,6], adversary-based approaches [32,11,33], and reconstruction-based approaches [34,8,9,35].”
Image-to-Image Translation: Methods and Applications (2021, methodology): “Image-to-cartoon \ \ [130], [102], [36], [132], [134], [140], [32], [37], [35], [34] [76], [152] \ [81] \ [31], [98] \ \ \”
Applications of Generative Adversarial Networks (GANs): An Updated Review (2019, background): “They proved that SPGAN has better qualify the generated images for domain adaptation Liu et al. [69] presented a novel and unified deep learning framework which is capable of learning domain-invariant representation from data across multiple domains.”
Domain Agnostic Learning with Disentangled Representations (2019, methodology): “Recent work attempts to disentangle features in the latent space of autoencoders with adversarial training (Cao et al., 2018; Liu et al., 2018b; Odena et al., 2017; Lee et al., 2018).”
DRIT++: Diverse Image-to-Image Translation via Disentangled Representations (2019, background): “Furthermore, several approaches (Choi et al. 2018; Liu et al. 2018) extend the conventional dual-domain I2I to general multidomain settings.”
Domain Fingerprints for No-Reference Image Quality Assessment (2019, methodology): “[23] proposes a model that is able to perform continuous cross-domain image translation and exhibits ability to learn and disentangle desirable latent representations.”
Federated Adversarial Domain Adaptation (2019, unknown): “To this end, recent work (Mathieu et al., 2016; Makhzani et al., 2016; Liu et al., 2018a; Odena et al., 2017) explores the learning of interpretable representations using generative adversarial networks (GANs) (Goodfellow et al., 2014) and variational autoencoders (VAEs) (Kingma & Welling, 2013).”
InfoGCN: Representation Learning for Human Skeleton-based Action Recognition (2022, background): “A large body of works have demonstrated that representation learning considerably influences the performance of machine learning tasks [2, 5, 13, 23, 29, 59, 61].”
Synthetic Data for Deep Learning (2019, background): “Further developments include asymmetric training for unsupervised domain adaptation [514], DistanceGAN for one-sided domain mapping [52], co-regularized alignment [334], cross-domain autoencoders [206], multisource domain adversarial networks [705], continuous cross-domain translation [367], face re”