Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wu, Yichen, Liu, Xu, Zhao, Chenxuan, Wu, Xinyu
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2509.18619
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912601314689024
author Wu, Yichen
Liu, Xu
Zhao, Chenxuan
Wu, Xinyu
author_facet Wu, Yichen
Liu, Xu
Zhao, Chenxuan
Wu, Xinyu
contents Inverting corrupted images into the latent space of diffusion models is challenging. Current methods, which encode an image into a single latent vector, struggle to balance structural fidelity with semantic accuracy, leading to reconstructions with semantic drift, such as blurred details or incorrect attributes. To overcome this, we introduce Prompt-Guided Dual Latent Steering (PDLS), a novel, training-free framework built upon Rectified Flow models for their stable inversion paths. PDLS decomposes the inversion process into two complementary streams: a structural path to preserve source integrity and a semantic path guided by a prompt. We formulate this dual guidance as an optimal control problem and derive a closed-form solution via a Linear Quadratic Regulator (LQR). This controller dynamically steers the generative trajectory at each step, preventing semantic drift while ensuring the preservation of fine detail without costly, per-image optimization. Extensive experiments on FFHQ-1K and ImageNet-1K under various inversion tasks, including Gaussian deblurring, motion deblurring, super-resolution and freeform inpainting, demonstrate that PDLS produces reconstructions that are both more faithful to the original image and better aligned with the semantic information than single-latent baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2509_18619
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Prompt-Guided Dual Latent Steering for Inversion Problems
Wu, Yichen
Liu, Xu
Zhao, Chenxuan
Wu, Xinyu
Computer Vision and Pattern Recognition
Inverting corrupted images into the latent space of diffusion models is challenging. Current methods, which encode an image into a single latent vector, struggle to balance structural fidelity with semantic accuracy, leading to reconstructions with semantic drift, such as blurred details or incorrect attributes. To overcome this, we introduce Prompt-Guided Dual Latent Steering (PDLS), a novel, training-free framework built upon Rectified Flow models for their stable inversion paths. PDLS decomposes the inversion process into two complementary streams: a structural path to preserve source integrity and a semantic path guided by a prompt. We formulate this dual guidance as an optimal control problem and derive a closed-form solution via a Linear Quadratic Regulator (LQR). This controller dynamically steers the generative trajectory at each step, preventing semantic drift while ensuring the preservation of fine detail without costly, per-image optimization. Extensive experiments on FFHQ-1K and ImageNet-1K under various inversion tasks, including Gaussian deblurring, motion deblurring, super-resolution and freeform inpainting, demonstrate that PDLS produces reconstructions that are both more faithful to the original image and better aligned with the semantic information than single-latent baselines.
title Prompt-Guided Dual Latent Steering for Inversion Problems
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.18619