{
  "abstract": "实时微型物体探测是低空无人飞行器(UAV)感知的关键任务。然而,由于成像高度、视角和复杂背景的变化,空中目标通常体积小、分布密集且视觉模糊,从而对检测精度和推理效率构成了巨大挑战。尽管RT-DETR框架提供了端到端的检测模式,但它对微小物体的反应较弱,特征融合过程中细微细节的退化,以及无人机场景中不稳定的边界箱回归。为应对这些限制,我们的FAMO-DETR被引入为一种频率感知的端到端探测器,专为通过无人机探测微小物体而设计。具体而言,通过离散的小波变换,将频率感知的Mamba编码器(FAME)设计用于将低频语义信息与高频边缘细节分离,从而通过Mamba序列建模,增强深度特征中的全局情境表示。随后开发出一种高频CARAFE模块(HF-CARAFE),将高频边缘引导和动态内核重新组装为特征融合,从而缓解传统升级操作引起的语义错位和边界模糊。最后,开发了NWD-CIoU损耗(NC-Loss),将归一化的Wasserstein距离的平滑距离约束与CIOU的几何重叠约束相结合,为微小物体定位提供了更稳定的优化信号。对VisDrone2019-DET和AI-TOD数据集的全面实验表明,FAMO-DETR实现了 $$AP_s$$ 在VisDrone2019-DET上,数值为22.1%,在AI-TOD上为AP值为23.5%,使基准配置分别提高了1.7%和2.3%,同时保持了134帧率的实时推理速度。",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": null,
  "author_total": 7,
  "authors": [
    {
      "name": "Sen Zhang"
    },
    {
      "name": "Chonghang Wang"
    },
    {
      "name": "Chenyang Li"
    },
    {
      "name": "Zhuangzhuang Zhou"
    },
    {
      "name": "Dengxi Chen"
    },
    {
      "name": "Shukui Bo"
    },
    {
      "name": "Rong Nie"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/fzamgkzpayrs2.bib",
  "cited_by": 0,
  "doi": "10.1038/s41598-026-73297-x",
  "fulltext_url": null,
  "id": "fzamgkzpayrs2",
  "issns": [
    "2045-2322"
  ],
  "landing_url": "https://www.nature.com/articles/s41598-026-73297-x",
  "language": null,
  "licence": "CC-BY-4.0",
  "licence_label": "CC BY 4.0",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0",
  "metadata_licence": null,
  "native_subjects": [
    [
      "openalex",
      "1707|Advanced Neural Network Applications"
    ]
  ],
  "openalex": "W7220679312",
  "pdf_url": "https://www.nature.com/articles/s41598-026-73297-x_reference.pdf",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-06",
  "publisher": "Springer Science and Business Media LLC",
  "ris": "https://research.draconitic.com/paper/fzamgkzpayrs2.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:17:22Z",
      "native_id": "10.1038/s41598-026-73297-x",
      "source": "crossref",
      "url": "https://www.nature.com/articles/s41598-026-73297-x"
    }
  ],
  "subjects": [],
  "title": "Frequency-aware Mamba and dynamic optimization DETR for real-time tiny-object detection in UAV aerial imagery",
  "title_text": "Frequency-aware Mamba and dynamic optimization DETR for real-time tiny-object detection in UAV aerial imagery",
  "type": "peer-reviewed",
  "updated_at": "2026-10-30T12:17:22Z",
  "url": "https://research.draconitic.com/paper/fzamgkzpayrs2",
  "venue": "Scientific Reports",
  "versions": []
}