Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codes
- 28 October 2022
- book chapter
- Published by Springer Nature
- p. 170-188
Abstract
No abstract availableKeywords
All Related Versions
This publication has 21 references indexed in Scilit:
- Towards a Neural Graphics Pipeline for Controllable Image GenerationComputer Graphics Forum, 2021
- Latent space manipulation for high-resolution medical image synthesis via the StyleGANZeitschrift für Medizinische Physik, 2020
- Analyzing and Improving the Image Quality of StyleGAN2020
- Non-Autoregressive Machine Translation with Latent Alignments2020
- Texture Mixer: A Network for Controllable Synthesis and Interpolation of Texture2019
- A Style-Based Generator Architecture for Generative Adversarial Networks2019
- Mask-Predict: Parallel Decoding of Conditional Masked Language Models2019
- The Unreasonable Effectiveness of Deep Features as a Perceptual Metric2018
- Generating Sentences from a Continuous Space2016
- Training Products of Experts by Minimizing Contrastive DivergenceNeural Computation, 2002