Latent Normalizing Flows for Many-to-Many Cross-Domain Mappings
This addresses the challenge of cross-domain mapping for tasks like image captioning and text-to-image synthesis, offering a novel approach that is incremental in improving representation learning.
The paper tackles the problem of learning joint representations for cross-domain tasks like image captioning by proposing a semi-supervised framework that models shared and domain-specific information separately, using invertible neural networks and normalizing flow-based priors to enable diverse many-to-many mappings, and demonstrates effectiveness on tasks such as image captioning and text-to-image synthesis.
Learned joint representations of images and text form the backbone of several important cross-domain tasks such as image captioning. Prior work mostly maps both domains into a common latent representation in a purely supervised fashion. This is rather restrictive, however, as the two domains follow distinct generative processes. Therefore, we propose a novel semi-supervised framework, which models shared information between domains and domain-specific information separately. The information shared between the domains is aligned with an invertible neural network. Our model integrates normalizing flow-based priors for the domain-specific information, which allows us to learn diverse many-to-many mappings between the two domains. We demonstrate the effectiveness of our model on diverse tasks, including image captioning and text-to-image synthesis.