From Diffusion to Rectified Flow: Rethinking Text-Based Segmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Qu, Zishen, Li, Xuesong, Gu, Haijian, Kang, Hongwei, Meng, Quan, Niu, Tianrui, Yang, Xin, Pan, Ruidong
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909017050185728
author Qu, Zishen
Li, Xuesong
Gu, Haijian
Kang, Hongwei
Meng, Quan
Niu, Tianrui
Yang, Xin
Pan, Ruidong
author_facet Qu, Zishen
Li, Xuesong
Gu, Haijian
Kang, Hongwei
Meng, Quan
Niu, Tianrui
Yang, Xin
Pan, Ruidong
contents Text-based image segmentation aims to delineate object boundaries within an image from text prompts, offering higher flexibility and broader application scope compared to traditional fixed-category segmentation tasks. Recent studies have shown that diffusion models (e.g., Stable Diffusion) can provide rich multimodal semantic features, leading to studies of using diffusion models as feature extractors for segmentation tasks. Such methods, however, inherit the generative natures of diffusion models that are harmful to discriminative segmentation tasks. In response, we propose RLFSeg, a novel framework that leverages Rectified Flow to learn direct mapping from the image to the segmentation mask within the latent space. The model is thus freed from the noise-denoise process and the need to optimize the time step of diffusion models, resulting in substantially better performance than previous diffusion-based methods, especially on zero-shot scenarios. By introducing label refinement and an Adaptive One-Step Sampling strategy, the model achieves higher accuracy even on a single inference step. The framework redirects a pretrained generative model to the discriminative segmentation task with zero modification to model structure, thus reveals promising application potential and significant research value.
format Preprint
id arxiv_https___arxiv_org_abs_2605_04590
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle From Diffusion to Rectified Flow: Rethinking Text-Based Segmentation
Qu, Zishen
Li, Xuesong
Gu, Haijian
Kang, Hongwei
Meng, Quan
Niu, Tianrui
Yang, Xin
Pan, Ruidong
Computer Vision and Pattern Recognition
Artificial Intelligence
Text-based image segmentation aims to delineate object boundaries within an image from text prompts, offering higher flexibility and broader application scope compared to traditional fixed-category segmentation tasks. Recent studies have shown that diffusion models (e.g., Stable Diffusion) can provide rich multimodal semantic features, leading to studies of using diffusion models as feature extractors for segmentation tasks. Such methods, however, inherit the generative natures of diffusion models that are harmful to discriminative segmentation tasks. In response, we propose RLFSeg, a novel framework that leverages Rectified Flow to learn direct mapping from the image to the segmentation mask within the latent space. The model is thus freed from the noise-denoise process and the need to optimize the time step of diffusion models, resulting in substantially better performance than previous diffusion-based methods, especially on zero-shot scenarios. By introducing label refinement and an Adaptive One-Step Sampling strategy, the model achieves higher accuracy even on a single inference step. The framework redirects a pretrained generative model to the discriminative segmentation task with zero modification to model structure, thus reveals promising application potential and significant research value.
title From Diffusion to Rectified Flow: Rethinking Text-Based Segmentation
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2605.04590