一个高效的基于YOLOv12的框架,用于检测极小规模的物体
A Chandrashekhar1, B Satyanarayana2, Rajani Reddy Gorrepati3
1Department of Mechanical Engineering, Faculty of Science and Technology, Icfai Foundation for Higher Education, Hyderabad, 501203, Telangana, India.
Scientific reports
|December 12, 2025
概括
这项研究介绍了一种高效的YOLOv12模型,用于从无人机上拍摄的空中图像中检测小物体. 新的区域注意力C2f (A2C2F) 模块和跨阶段部分与内核大小2 (C3K2) 模块增强了功能学习,并降低了用于稳健检测的计算成本.
科学领域:
- 计算机视觉 计算机视觉
- 人工智能的人工智能
- 遥感 遥感 遥感 遥感
背景情况:
- 在无人机的空中图像中对象检测面临着小物体和极小物体的挑战.
- 现有的YOLO变种在无人机捕获的数据中难以检测微小的物体.
研究的目的:
- 提出一种高效的YOLOv12模型,用于在空中成像中增强小型和极小物体检测.
- 引入新的模块和策略,以提高特征学习和检测准确度.
主要方法:
- 开发了区域注意力C2f (A2C2F) 模块,将多头MLP与局部区域注意力融合在一起.
- 集成了轻量级的交叉阶段部分与内核大小2 (C3K2) 模块,以减少计算复杂性.
- 实施了A2C2F块的多尺度融合策略,以保持高分辨率的特征.
- 设计了一种新的检测头,分类和回归分支脱,注意力引导的特征融合.
主要成果:
- 拟议的YOLOv12模型有效检测小于3-5像素的物体,性能优于传统的YOLO变体.
- 获得的精度:69.1%,回忆:48.5%,F1得分:56.99%,mAP@50:58.8%,mAP@0.5:0.95:40.9%在VisDrone数据集上.
- 在A100 GPU上以59.1M参数和198.6 GFLOPs的速度在~40 FPS的速度上演示了高效的推断.
- 在较低输入分辨率 (640x640) 的情况下,性能优于双阶段和基模型.
结论:
- 拟议的YOLOv12框架为基于无人机的空中图像中的小型和极小物体检测提供了强大的和计算效率高的解决方案.
- 新的A2C2F和C3K2模块,以及多尺度融合和检测头,显著提高了检测性能和推断速度.
- 该模型适用于实时无人机边缘部署,因为其在具有挑战性的检测场景中的效率和高精度.
相关概念视频
01:10Detection of Black Holes
2.5K
Although black holes were theoretically postulated in the 1920s, they remained outside the domain of observational astronomy until the 1970s.
Their closest cousins are neutron stars, which are composed almost entirely of neutrons packed against each other, making them extremely dense. A neutron star has the same mass as the Sun but its diameter is only a few kilometers. Therefore, the escape velocity from their surface is close to the speed of light.
Not until the 1960s, when the first neutron...
Their closest cousins are neutron stars, which are composed almost entirely of neutrons packed against each other, making them extremely dense. A neutron star has the same mass as the Sun but its diameter is only a few kilometers. Therefore, the escape velocity from their surface is close to the speed of light.
Not until the 1960s, when the first neutron...
2.5K
01:05Difference from Background: Limit of Detection
8.0K
The limit of detection (LOD) is the smallest amount of analyte that can be distinguished from the background noise. The LOD value corresponds to the concentration at which the analyte signal is three times larger than the standard deviation of the blank signal. Below this value, the analyte signal cannot be differentiated from the background noise. It is calculated by dividing the calibration slope by 3 times the standard deviation of the blank signals.
The LOD indicates the presence or absence...
The LOD indicates the presence or absence...
8.0K
01:29Two-Dimensional Microscopy in Microbiology
1.0K
Two-dimensional (2D) microscopy encompasses a range of optical techniques that capture images within a single focal plane, offering detailed representations of microscopic structures. These techniques are essential in biological and medical research, enabling the visualization of cellular and subcellular structures with different levels of contrast and specificity.There are several major types of 2D microscopy, each with strengths and applications.Bright-Field MicroscopyBright-field microscopy...
1.0K
02:07Improving Translational Accuracy
14.0K
Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...
14.0K
02:07Improving Translational Accuracy
3.5K
3.5K
01:07Electron Microscope Tomography and Single-particle Reconstruction
2.8K
Transmission electron microscopy (TEM) can be used to determine the 3D structure of biological samples with the help of techniques such as electron microscope tomography and single-particle reconstruction. While single-particle reconstruction can examine macromolecules and macromolecular complexes in vitro conditions only, tomography permits the study of cell components or small cells in vivo.
Electron Tomography
Electron tomography can be performed either in TEM or STEM (scanning transmission...
Electron Tomography
Electron tomography can be performed either in TEM or STEM (scanning transmission...
2.8K