arrow
返回

Text-to-Vector Generation with Neural Path Representation

delete2024-07-19
delete0
delete
OA
AI
P
Peiying Zhang
N
Nanxuan Zhao
廖静 (Jing Liao) *
DOI:10.1145/3658204delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
Vector graphics are widely used in digital art and highly favored by designers due to their scalability and layer-wise properties. However, the process of creating and editing vector graphics requires creativity and design expertise, making it a time-consuming task. Recent advancements in text-to-vector (T2V) generation have aimed to make this process more accessible. However, existing T2V methods directly optimize control points of vector graphics paths, often resulting in intersecting or jagged paths due to the lack of geometry constraints. To overcome these limitations, we propose a novel neural path representation by designing a dual-branch Variational Autoencoder (VAE) that learns the path latent space from both sequence and image modalities. By optimizing the combination of neural paths, we can incorporate geometric constraints while preserving expressivity in generated SVGs. Furthermore, we introduce a two-stage path optimization method to improve the visual and topological quality of generated SVGs. In the first stage, a pre-trained text-to-image diffusion model guides the initial generation of complex vector graphics through the Variational Score Distillation (VSD) process. In the second stage, we refine the graphics using a layer-wise image vectorization strategy to achieve clearer elements and structure. We demonstrate the effectiveness of our method through extensive experiments and showcase various applications. The project page is https://intchous.github.io/T2V-NPR.
Keyword:
Vector Graphics
SVG
Diffusion Model
Text-Guided Generation

期刊

ACM Transactions on Graphics 封面图
ACM Transactions on Graphics
IF:
9.5
论文数:
4.7K
被引数:
3.6W

机构

A
adobe systems inc.
学者数:
273
论文数: 305
被引数: 0
C
City University of Hong Kong
学者数:
2.3W
论文数: 3.0W
被引数: 6.1W
引用论文

引用论文

Perception-Driven Semi-Structured Boundary Vectorization
err2018-07-30
err23
errOAAI
errHoshyari, Shayan; Alberto Dominici, Edoardo; Sheffer, Alla; Carr, Nathan; Wang, Zhaowen; Ceylan, Duygu; Shen, I-Chao
err分享
err收藏
Examination of the Functional Relationship between PD-L1 DNA Methylation and mRNA Expression in Non-Small-Cell Lung Cancer
err2023-03-22
err0
errOAAI
errTrine V. Larsen; Nina Dybdal; Tina F. Daugaard; Johanne Lade-Keller; Lin Lin; Boe S. Sorensen; Anders L. Nielsen
err分享
err收藏
CLIPasso: Semantically-Aware Object Sketching
err2022-07-22
err73
PREAI
errVinker, Yael; Pajouheshgar, Ehsan; Bo, Jessica Y.; Bachmann, Roman Christian; Bermano, Amit Haim; Cohen-Or, Daniel; Zamir, Amir; Shamir, Ariel
err分享
err收藏
err分享
err收藏
The comorbidity and co-medication profile of patients with progressive supranuclear palsy
err2023-10-06
err0
errOAAI
errStephan Greten; Florian Wegner; Ida Jensen; Lea Krey; Sophia Rogozinski; Meret Fehring; Johanne Heine; Johanna Doll-Lee; Monika Pötter-Nerger; Molly Zeitzschel; Keno Hagena; David J. Pedrosa; Carsten Eggers; Katrin Bürk; Claudia Trenkwalder; Inga Claus; Tobias Warnecke; Patrick Süß; Jürgen Winkler; Doreen Gruber; Florin Gandor; Daniela Berg; Steffen Paschen; Joseph Classen; Elmar H. Pinkhardt; Jan Kassubek; Wolfgang H. Jost; Lars Tönges; Andrea A. Kühn; Johannes Schwarz; Oliver Peters; Eman Dashti; Josef Priller; Eike J. Spruth; Patricia Krause; Annika Spottke; Anja Schneider; Aline Beyle; Okka Kimmich; Markus Donix; Robert Haussmann; Moritz Brandt; Elisabeth Dinter; Jens Wiltfang; Björn H. Schott; Inga Zerr; Mathias Bähr; Katharina Buerger; Daniel Janowitz; Robert Perneczky; Boris-Stephan Rauchmann; Endy Weidinger; Johannes Levin; Sabrina Katzdobler; Emrah Düzel; Wenzel Glanz; Stefan Teipel; Ingo Kilimann; Johannes Prudlo; Thomas Gasser; Kathrin Brockmann; Daniel C. Hoffmann; Thomas Klockgether; Olaf Krause; Johannes Heck; Günter U. Höglinger; Martin Klietz
err分享
err收藏
err分享
err收藏
学者 查看更多内容