FOUND: Fourier-based von Mises Distribution for Robust Single Domain Generalization in Object Detection

TL;DR

This paper introduces a Fourier-based von Mises distribution framework integrated with CLIP to improve single domain generalization in object detection by modeling feature distributions and augmenting frequency components for robustness.

Contribution

It presents a novel approach combining vMF distribution modeling and Fourier augmentation within a CLIP-guided pipeline for enhanced domain generalization in object detection.

Findings

Outperforms state-of-the-art methods on weather-driving benchmarks.

Effectively models domain-invariant features using vMF distribution.

Improves robustness through frequency-domain augmentation.

Abstract

Single Domain Generalization (SDG) for object detection aims to train a model on a single source domain that can generalize effectively to unseen target domains. While recent methods like CLIP-based semantic augmentation have shown promise, they often overlook the underlying structure of feature distributions and frequency-domain characteristics that are critical for robustness. In this paper, we propose a novel framework that enhances SDG object detection by integrating the von Mises-Fisher (vMF) distribution and Fourier transformation into a CLIP-guided pipeline. Specifically, we model the directional features of object representations using vMF to better capture domain-invariant semantic structures in the embedding space. Additionally, we introduce a Fourier-based augmentation strategy that perturbs amplitude and phase components to simulate domain shifts in the frequency domain, further improving feature robustness. Our method not only preserves the semantic alignment benefits of CLIP but also enriches feature diversity and structural consistency across domains. Extensive experiments on the diverse weather-driving benchmark demonstrate that our approach outperforms the existing state-of-the-art method.