AssetDropper: Asset Extraction via Diffusion Models with Reward-Driven Optimization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Lanjiong, Zhao, Guanhua, Zhu, Lingting, Cai, Zeyu, Yu, Lequan, Zhang, Jian, Wang, Zeyu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912419834494976
author Li, Lanjiong
Zhao, Guanhua
Zhu, Lingting
Cai, Zeyu
Yu, Lequan
Zhang, Jian
Wang, Zeyu
author_facet Li, Lanjiong
Zhao, Guanhua
Zhu, Lingting
Cai, Zeyu
Yu, Lequan
Zhang, Jian
Wang, Zeyu
contents Recent research on generative models has primarily focused on creating product-ready visual outputs; however, designers often favor access to standardized asset libraries, a domain that has yet to be significantly enhanced by generative capabilities. Although open-world scenes provide ample raw materials for designers, efficiently extracting high-quality, standardized assets remains a challenge. To address this, we introduce AssetDropper, the first framework designed to extract assets from reference images, providing artists with an open-world asset palette. Our model adeptly extracts a front view of selected subjects from input images, effectively handling complex scenarios such as perspective distortion and subject occlusion. We establish a synthetic dataset of more than 200,000 image-subject pairs and a real-world benchmark with thousands more for evaluation, facilitating the exploration of future research in downstream tasks. Furthermore, to ensure precise asset extraction that aligns well with the image prompts, we employ a pre-trained reward model to fulfill a closed-loop with feedback. We design the reward model to perform an inverse task that pastes the extracted assets back into the reference sources, which assists training with additional consistency and mitigates hallucination. Extensive experiments show that, with the aid of reward-driven optimization, AssetDropper achieves the state-of-the-art results in asset extraction. Project page: AssetDropper.github.io.
format Preprint
id arxiv_https___arxiv_org_abs_2506_07738
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle AssetDropper: Asset Extraction via Diffusion Models with Reward-Driven Optimization
Li, Lanjiong
Zhao, Guanhua
Zhu, Lingting
Cai, Zeyu
Yu, Lequan
Zhang, Jian
Wang, Zeyu
Computer Vision and Pattern Recognition
Recent research on generative models has primarily focused on creating product-ready visual outputs; however, designers often favor access to standardized asset libraries, a domain that has yet to be significantly enhanced by generative capabilities. Although open-world scenes provide ample raw materials for designers, efficiently extracting high-quality, standardized assets remains a challenge. To address this, we introduce AssetDropper, the first framework designed to extract assets from reference images, providing artists with an open-world asset palette. Our model adeptly extracts a front view of selected subjects from input images, effectively handling complex scenarios such as perspective distortion and subject occlusion. We establish a synthetic dataset of more than 200,000 image-subject pairs and a real-world benchmark with thousands more for evaluation, facilitating the exploration of future research in downstream tasks. Furthermore, to ensure precise asset extraction that aligns well with the image prompts, we employ a pre-trained reward model to fulfill a closed-loop with feedback. We design the reward model to perform an inverse task that pastes the extracted assets back into the reference sources, which assists training with additional consistency and mitigates hallucination. Extensive experiments show that, with the aid of reward-driven optimization, AssetDropper achieves the state-of-the-art results in asset extraction. Project page: AssetDropper.github.io.
title AssetDropper: Asset Extraction via Diffusion Models with Reward-Driven Optimization
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2506.07738