FlowDIS: Language-Guided Dichotomous Image Segmentation with Flow Matching

TL;DR

FlowDIS introduces a novel flow matching-based approach for language-guided dichotomous image segmentation, significantly improving accuracy and controllability over existing methods.

Contribution

The paper presents FlowDIS, a new flow matching framework with position-aware training for precise, language-guided image segmentation, outperforming prior state-of-the-art methods.

Findings

FlowDIS achieves 5.5% higher $F_{eta}^{ ext{w}}$ measure than previous methods.

FlowDIS reduces MAE by 43% compared to the best prior DIS approach.

FlowDIS demonstrates strong controllability and fine-grained segmentation with language prompts.

Abstract

Accurate image segmentation is essential for modern computer vision applications such as image editing, autonomous driving, and medical image analysis. In recent years, Dichotomous Image Segmentation (DIS) has become a standard task for training and evaluating highly accurate segmentation models. Existing DIS approaches often fail to preserve fine-grained details or fully capture the semantic structure of the foreground. To address these challenges, we present FlowDIS, a novel dichotomous image segmentation method built on the flow matching framework, which learns a time-dependent vector field to transport the image distribution to the corresponding mask distribution, optionally conditioned on a text prompt. Moreover, with our Position-Aware Instance Pairing (PAIP) training strategy, FlowDIS offers strong controllability through text prompts, enabling precise, pixel-level object segmentation. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art approaches both with and without language guidance. Compared with the best prior DIS method, FlowDIS achieves a 5.5% higher $F_{\beta}^{\omega}$ measure and 43% lower MAE ($\mathcal{M}$) on the DIS-TE test set. The code is available at: https://github.com/Picsart-AI-Research/FlowDIS