Cover Image

Meissonic: Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis

Saved in:
Published in:arXiv.org (Dec 5, 2024), p. n/a
Main Author: Bai, Jinbin
Other Authors: Tian Ye, Chow, Wei, Song, Enxin, Qing-Guo, Chen, Li, Xiangtai, Dong, Zhen, Zhu, Lei, Shuicheng Yan
Published:
Cornell University Library, arXiv.org
Subjects:
Diffusion rate
Image compression
Image resolution
Image quality
Image enhancement
Image processing
Synthesis
High resolution
Online Access:Citation/Abstract
Full text outside of ProQuest
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract:We present Meissonic, which elevates non-autoregressive masked image modeling (MIM) text-to-image to a level comparable with state-of-the-art diffusion models like SDXL. By incorporating a comprehensive suite of architectural innovations, advanced positional encoding strategies, and optimized sampling conditions, Meissonic substantially improves MIM's performance and efficiency. Additionally, we leverage high-quality training data, integrate micro-conditions informed by human preference scores, and employ feature compression layers to further enhance image fidelity and resolution. Our model not only matches but often exceeds the performance of existing models like SDXL in generating high-quality, high-resolution images. Extensive experiments validate Meissonic's capabilities, demonstrating its potential as a new standard in text-to-image synthesis. We release a model checkpoint capable of producing \(1024 \times 1024\) resolution images.
ISSN:2331-8422
Source:Engineering Database