Representative Feature Extraction During Diffusion Process for Sketch Extraction with One Example

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yun, Kwan, Kim, Youngseo, Seo, Kwanggyoon, Seo, Chang Wook, Noh, Junyong
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929204059176960
author Yun, Kwan
Kim, Youngseo
Seo, Kwanggyoon
Seo, Chang Wook
Noh, Junyong
author_facet Yun, Kwan
Kim, Youngseo
Seo, Kwanggyoon
Seo, Chang Wook
Noh, Junyong
contents We introduce DiffSketch, a method for generating a variety of stylized sketches from images. Our approach focuses on selecting representative features from the rich semantics of deep features within a pretrained diffusion model. This novel sketch generation method can be trained with one manual drawing. Furthermore, efficient sketch extraction is ensured by distilling a trained generator into a streamlined extractor. We select denoising diffusion features through analysis and integrate these selected features with VAE features to produce sketches. Additionally, we propose a sampling scheme for training models using a conditional generative approach. Through a series of comparisons, we verify that distilled DiffSketch not only outperforms existing state-of-the-art sketch extraction methods but also surpasses diffusion-based stylization methods in the task of extracting sketches.
format Preprint
id arxiv_https___arxiv_org_abs_2401_04362
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Representative Feature Extraction During Diffusion Process for Sketch Extraction with One Example
Yun, Kwan
Kim, Youngseo
Seo, Kwanggyoon
Seo, Chang Wook
Noh, Junyong
Computer Vision and Pattern Recognition
Artificial Intelligence
Graphics
68T01
I.4.9
We introduce DiffSketch, a method for generating a variety of stylized sketches from images. Our approach focuses on selecting representative features from the rich semantics of deep features within a pretrained diffusion model. This novel sketch generation method can be trained with one manual drawing. Furthermore, efficient sketch extraction is ensured by distilling a trained generator into a streamlined extractor. We select denoising diffusion features through analysis and integrate these selected features with VAE features to produce sketches. Additionally, we propose a sampling scheme for training models using a conditional generative approach. Through a series of comparisons, we verify that distilled DiffSketch not only outperforms existing state-of-the-art sketch extraction methods but also surpasses diffusion-based stylization methods in the task of extracting sketches.
title Representative Feature Extraction During Diffusion Process for Sketch Extraction with One Example
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Graphics
68T01
I.4.9
url https://arxiv.org/abs/2401.04362