From A to Z: Supervised Transfer of Style and Content Using Deep Neural Network Generators
arXiv:1603.02003
Abstract
We propose a new neural network architecture for solving single-image analogies - the generation of an entire set of stylistically similar images from just a single input image. Solving this problem requires separating image style from content. Our network is a modified variational autoencoder (VAE) that supports supervised training of single-image analogies and in-network evaluation of outputs with a structured similarity objective that captures pixel covariances. On the challenging task of generating a 62-letter font from a single example letter we produce images with 22.4% lower dissimilarity to the ground truth than state-of-the-art.
References in corpus (4)
Cited by in corpus (14)
- Generalizing Across Domains via Cross-Gradient Training
- Artistic Glyph Image Synthesis via One-Stage Few-Shot Learning
- Separating Style and Content for Generalized Style Transfer
- DG-Font: Deformable Generative Networks for Unsupervised Font Generation
- A Unified Framework for Generalizable Style Transfer: Style and Content Separation
- DiVa-360: The Dynamic Visual Dataset for Immersive Neural Fields
- Handwritten Chinese Font Generation with Collaborative Stroke Refinement
- Typeface Completion with Generative Adversarial Networks
- Learning Perceptual Manifold of Fonts
- Natural Scene Text Editing Based on AI
- Semantic Vector Spaces for Broadening Consideration of Consequences
- Font Completion and Manipulation by Cycling Between Multi-Modality Representations
- Generative Adversarial Networks with Inverse Transformation Unit
- Few-Shot Font Generation with Deep Metric Learning