DeFloMat: Detection with Flow Matching for Stable and Efficient Generative Object Localization

TL;DR

DeFloMat introduces a deterministic flow matching approach for object detection that significantly improves speed and accuracy, making diffusion-based methods practical for clinical applications like MRI analysis.

Contribution

It replaces slow stochastic diffusion processes with a direct, deterministic flow derived from optimal transport, enabling rapid inference with fewer steps.

Findings

Achieves 43.32% AP_{10:50} in 3 steps, outperforming previous methods.

Reduces inference steps from 4 to 3, improving performance by 1.4x.

Enhances localization stability and recall in few-step regimes.

Abstract

We propose DeFloMat (Detection with Flow Matching), a novel generative object detection framework that addresses the critical latency bottleneck of diffusion-based detectors, such as DiffusionDet, by integrating Conditional Flow Matching (CFM). Diffusion models achieve high accuracy by formulating detection as a multi-step stochastic denoising process, but their reliance on numerous sampling steps ($T \gg 60$) makes them impractical for time-sensitive clinical applications like Crohn's Disease detection in Magnetic Resonance Enterography (MRE). DeFloMat replaces this slow stochastic path with a highly direct, deterministic flow field derived from Conditional Optimal Transport (OT) theory, specifically approximating the Rectified Flow. This shift enables fast inference via a simple Ordinary Differential Equation (ODE) solver. We demonstrate the superiority of DeFloMat on a challenging MRE clinical dataset. Crucially, DeFloMat achieves state-of-the-art accuracy ($43.32\% \text{ } AP_{10:50}$) in only $3$ inference steps, which represents a $1.4\times$ performance improvement over DiffusionDet's maximum converged performance ($31.03\% \text{ } AP_{10:50}$ at $4$ steps). Furthermore, our deterministic flow significantly enhances localization characteristics, yielding superior Recall and stability in the few-step regime. DeFloMat resolves the trade-off between generative accuracy and clinical efficiency, setting a new standard for stable and rapid object localization.