Symmetrical Flow Matching: Unified Image Generation, Segmentation, and Classification with Score-Based Generative Models

TL;DR

SymmFlow is a unified model leveraging symmetric flow matching for high-quality image generation, segmentation, and classification, achieving state-of-the-art results with efficient one-step inference and flexible conditioning.

Contribution

It introduces SymmFlow, a novel symmetric flow matching framework that unifies multiple tasks and improves efficiency and flexibility over previous models.

Findings

Achieves state-of-the-art FID scores on CelebAMask-HQ and COCO-Stuff.

Supports one-step semantic segmentation and classification.

Demonstrates competitive performance across multiple benchmarks.

Abstract

Flow Matching has emerged as a powerful framework for learning continuous transformations between distributions, enabling high-fidelity generative modeling. This work introduces Symmetrical Flow Matching (SymmFlow), a new formulation that unifies semantic segmentation, classification, and image generation within a single model. Using a symmetric learning objective, SymmFlow models forward and reverse transformations jointly, ensuring bi-directional consistency, while preserving sufficient entropy for generative diversity. A new training objective is introduced to explicitly retain semantic information across flows, featuring efficient sampling while preserving semantic structure, allowing for one-step segmentation and classification without iterative refinement. Unlike previous approaches that impose strict one-to-one mapping between masks and images, SymmFlow generalizes to flexible conditioning, supporting both pixel-level and image-level class labels. Experimental results on various benchmarks demonstrate that SymmFlow achieves state-of-the-art performance on semantic image synthesis, obtaining FID scores of 11.9 on CelebAMask-HQ and 7.0 on COCO-Stuff with only 25 inference steps. Additionally, it delivers competitive results on semantic segmentation and shows promising capabilities in classification tasks.