Adversarial Learning of Disentangled and Generalizable Representations for Visual Attributes
This work addresses the need for more flexible and generalizable models in image-to-image translation, particularly for tasks like multi-attribute manipulation, which is incremental as it builds on existing adversarial and autoencoder methods.
The paper tackles the problem of rigid image-to-image translation models that lack semantic latent structure and rely on binary domain labels, proposing a novel adversarial learning method that disentangles and generalizes latent representations for attributes. The result is a method that outperforms state-of-the-art approaches in intensity-preserving multi-attribute transfer and synthesis on datasets like MultiPIE, RaFD, and BU-3DFE.
Recently, a multitude of methods for image-to-image translation have demonstrated impressive results on problems such as multi-domain or multi-attribute transfer. The vast majority of such works leverages the strengths of adversarial learning and deep convolutional autoencoders to achieve realistic results by well-capturing the target data distribution. Nevertheless, the most prominent representatives of this class of methods do not facilitate semantic structure in the latent space, and usually rely on binary domain labels for test-time transfer. This leads to rigid models, unable to capture the variance of each domain label. In this light, we propose a novel adversarial learning method that (i) facilitates the emergence of latent structure by semantically disentangling sources of variation, and (ii) encourages learning generalizable, continuous, and transferable latent codes that enable flexible attribute mixing. This is achieved by introducing a novel loss function that encourages representations to result in uniformly distributed class posteriors for disentangled attributes. In tandem with an algorithm for inducing generalizable properties, the resulting representations can be utilized for a variety of tasks such as intensity-preserving multi-attribute image translation and synthesis, without requiring labelled test data. We demonstrate the merits of the proposed method by a set of qualitative and quantitative experiments on popular databases such as MultiPIE, RaFD, and BU-3DFE, where our method outperforms other, state-of-the-art methods in tasks such as intensity-preserving multi-attribute transfer and synthesis.