ParallelEdits: Efficient Multi-Aspect Text-Driven Image Editing with Attention Grouping

Text-driven image synthesis has made significant advancements with the development of diffusion models, transforming how visual content is generated from text prompts. Despite these advances, text-driven image editing, a key area in computer graphics, faces unique challenges. A major challenge is ma...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2024-11
Main Authors: Huang, Mingzhen, Cai, Jialing, Jia, Shan, Lokhande, Vishnu Suresh, Lyu, Siwei
Format: Article
Language:English
Subjects:
Computer graphics
Datasets
Editing
Multitasking
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Text-driven image synthesis has made significant advancements with the development of diffusion models, transforming how visual content is generated from text prompts. Despite these advances, text-driven image editing, a key area in computer graphics, faces unique challenges. A major challenge is making simultaneous edits across multiple objects or attributes. Applying these methods sequentially for multi-attribute edits increases computational demands and efficiency losses. In this paper, we address these challenges with significant contributions. Our main contribution is the development of ParallelEdits, a method that seamlessly manages simultaneous edits across multiple attributes. In contrast to previous approaches, ParallelEdits not only preserves the quality of single attribute edits but also significantly improves the performance of multitasking edits. This is achieved through innovative attention distribution mechanism and multi-branch design that operates across several processing heads. Additionally, we introduce the PIE-Bench++ dataset, an expansion of the original PIE-Bench dataset, to better support evaluating image-editing tasks involving multiple objects and attributes simultaneously. This dataset is a benchmark for evaluating text-driven image editing methods in multifaceted scenarios.
ISSN:2331-8422