用扩散模型分步对齐仿真与真实无人机图像，提升小目标检测精度。
Coarse-to-Fine Hierarchical Alignment for UAV-based Human Detection using Diffusion Models
- 分三阶段：先调整体风格，再细化局部细节，最后剔除不真实人体实例。
- 在语义无人机数据集上，mAP50最高提升14.1个百分点。
- 适合需要高精度无人机行人检测且标注困难的场景。
训练目标检测器需大量特定任务标注数据，但在无人机行检测中，由于目标分布不断变化且标注图像稀缺，该要求难以实现。为此，采用合成模拟器生成带标注数据，虽成本低，但合成与真实图像间存在域差距，影响模型在真实场景的应用效果。为此，本文提出粗到细层次对齐（CFHA）框架，基于扩散模型将合成数据转换为更贴近真实分布的数据，同时保留原始标注。该方法显式分离全局风格与局部内容的域差异，通过三个模块解决：(1) 全局风格迁移——使用少量真实参考图像，通过扩散模型对齐合成图像的颜色、光照和纹理统计特征；(2) 局部精修——采用超分辨率扩散模型增强小目标（如人体）的细微真实感细节，保持形状与边界完整性；(3) 虚构消除——过滤视觉属性不符合真实数据的人体实例，使外观更贴近目标分布。在公开的无人机仿真到真实检测基准上进行大量实验，结果表明，本方法显著优于未经转换的基线模型。具体而言，在Semantic-Drone数据集上，mAP50最高提升14.1。消融实验证明全局与局部阶段具有互补性，且层级对齐至关重要。代码已开源。
原文摘要 · Abstract (English)
Training object detectors demands extensive, task-specific annotations, yet this requirement becomes impractical in UAV-based human detection due to constantly shifting target distributions and the scarcity of labeled images. As a remedy, synthetic simulators are adopted to generate annotated data, with a low annotation cost. However, the domain gap between synthetic and real images hinders the model from being effectively applied to the target domain. Accordingly, we introduce Coarse-to-Fine Hierarchical Alignment (CFHA), a three-stage diffusion-based framework designed to transform synthetic data for UAV-based human detection, narrowing the domain gap while preserving the original synthetic labels. CFHA explicitly decouples global style and local content domain discrepancies and bridges those gaps using three modules: (1) Global Style Transfer -- a diffusion model aligns color, illumination, and texture statistics of synthetic images to the realistic style, using only a small real reference set; (2) Local Refinement -- a super-resolution diffusion model is used to facilitate fine-grained and photorealistic details for the small objects, such as human instances, preserving shape and boundary integrity; (3) Hallucination Removal -- a module that filters out human instances whose visual attributes do not align with real-world data to make the human appearance closer to the target distribution. Extensive experiments on public UAV Sim2Real detection benchmarks demonstrate that our methods significantly improve the detection accuracy compared to the non-transformed baselines. Specifically, our method achieves up to $+14.1$ improvement of mAP50 on Semantic-Drone benchmark. Ablation studies confirm the complementary roles of the global and local stages and highlight the importance of hierarchical alignment. The code is released at \href{https://github.com/liwd190019/CFHA}{this url}.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。