GLANCEGuided Language Through Autoregression Establishing Natural and Classifier-Free Editing

Title

GLANCEGuided Language Through Autoregression Establishing Natural and Classifier-Free Editing

Creator

Joshi D.; Tank A.; Pramila R.M.

Description

In this study, researchers aimed to simplify text conversion into images using the latest text-to-image generation methods. While these methods have improved the quality and relevance of generated images, certain crucial questions remained unanswered, limiting their practicality and overall quality. To address these issues, the researchers introduced a novel text-to-image method. This method allows for better control of the scene depicted in the image through text, enhances the tokenization process by incorporating specific knowledge about key image regions such as faces and important objects, and provides guidance to the transformer model without needing a classifier. The outcome of this work was a model that achieved state-of-the-art results in terms of image quality and human evaluation, enabling the generation of high-fidelity 512?512-pixel images. Moreover, this method introduced new capabilities, including scene editing, text editing with reference scenes, handling out-of-distribution text prompts, and generating story illustrations. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

Source

Lecture Notes in Networks and Systems, Vol-922 LNNS, pp. 93-102.

Date

2024-01-01

Publisher

Springer Science and Business Media Deutschland GmbH

Subject

Autoregression; Classifier-free guidance; Diffusion models; Generative adversarial networks; Guided diffusion; Image inpainting; Image tokenization; Image-to-Image generation; Text-to-image generation

Coverage

Joshi D., Christ University, Bangalore, Indore, India; Tank A., Christ University, Bangalore, Indore, India; Pramila R.M., Christ University, Bangalore, Indore, India

Rights

Restricted Access

Relation

ISSN: 23673370; ISBN: 978-981970974-8

Format

Online

Language

English

Type

Conference paper

Identifier

https://doi.org/10.1007/978-981-97-0975-5_8

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85197240594&doi=10.1007%2f978-981-97-0975-5_8&partnerID=40&md5=d1bf9533386ce7d06aaab9a8e9411f09