Feedforward 3D Editing Learns from Semantic-Part Transformation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Weng, Jiawei, Zhang, Saining, Diao, Zhenxin, Li, Peishuo, Zhang, Henghaofan, Chen, Junhao, Zhao, Hao
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913168039608320
author Weng, Jiawei
Zhang, Saining
Diao, Zhenxin
Li, Peishuo
Zhang, Henghaofan
Chen, Junhao
Zhao, Hao
author_facet Weng, Jiawei
Zhang, Saining
Diao, Zhenxin
Li, Peishuo
Zhang, Henghaofan
Chen, Junhao
Zhao, Hao
contents 3D editing is a fundamental capability for scalable 3D content creation. While image editing has rapidly evolved toward large-scale feedforward generative paradigms, 3D AI generation remains dominated by training-free editing pipelines. A central challenge of feedforward 3D editing lies in the lack of high-quality paired supervision. Editable 3D assets require simultaneous preservation of geometry, multi-view consistency, structural coherence, and localized edit controllability. Existing 3D editing datasets often rely on independently generated assets, image-mediated reconstruction or narrow edit taxonomies, leading to inaccurate localization, weak preservation, blurred edit boundaries, and limited semantic consistency. In this work, we introduce a new perspective: scalable feedforward 3D editing should be learned from semantic-part transformations. Based on this insight, we propose Pxform, a high-quality 3D editing dataset with over 100K consistent before/after editing pairs across seven edit types. Instead of treating objects as unstructured shapes, our pipeline grounds edits directly in semantic 3D parts. Built upon Pxform, we further propose PartFlow, a feedforward 3D editing network that injects source-aware latent control into pretrained 3D generative priors. PartFlow introduces mask-aware velocity preservation and render-space consistency supervision to jointly improve edit fidelity and source preservation, while requiring no 3D edit mask during inference. Extensive experiments demonstrate that high-quality semantic-part supervision substantially improves scalable 3D editing, enabling PartFlow to achieve state-of-the-art performance on both geometric and appearance editing benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2605_27351
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Feedforward 3D Editing Learns from Semantic-Part Transformation
Weng, Jiawei
Zhang, Saining
Diao, Zhenxin
Li, Peishuo
Zhang, Henghaofan
Chen, Junhao
Zhao, Hao
Computer Vision and Pattern Recognition
3D editing is a fundamental capability for scalable 3D content creation. While image editing has rapidly evolved toward large-scale feedforward generative paradigms, 3D AI generation remains dominated by training-free editing pipelines. A central challenge of feedforward 3D editing lies in the lack of high-quality paired supervision. Editable 3D assets require simultaneous preservation of geometry, multi-view consistency, structural coherence, and localized edit controllability. Existing 3D editing datasets often rely on independently generated assets, image-mediated reconstruction or narrow edit taxonomies, leading to inaccurate localization, weak preservation, blurred edit boundaries, and limited semantic consistency. In this work, we introduce a new perspective: scalable feedforward 3D editing should be learned from semantic-part transformations. Based on this insight, we propose Pxform, a high-quality 3D editing dataset with over 100K consistent before/after editing pairs across seven edit types. Instead of treating objects as unstructured shapes, our pipeline grounds edits directly in semantic 3D parts. Built upon Pxform, we further propose PartFlow, a feedforward 3D editing network that injects source-aware latent control into pretrained 3D generative priors. PartFlow introduces mask-aware velocity preservation and render-space consistency supervision to jointly improve edit fidelity and source preservation, while requiring no 3D edit mask during inference. Extensive experiments demonstrate that high-quality semantic-part supervision substantially improves scalable 3D editing, enabling PartFlow to achieve state-of-the-art performance on both geometric and appearance editing benchmarks.
title Feedforward 3D Editing Learns from Semantic-Part Transformation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.27351