期刊文献+

基于改进YOLOv7⁃tiny的户外行人检测算法研究

Research on outdoor pedestrian detection algorithm based on improved YOLOv7⁃tiny
下载PDF
导出
摘要 在行人检测的高密度交通场景中,检测算法通常会漏掉被遮挡和远处的模糊行人,同时无法兼顾检测的精度和速度。针对这些问题,基于YOLOv7⁃tiny提出一种改进的户外行人检测算法。该算法引入SENet注意力机制,抑制了不相关的信息,以此提高特征图表达信息的能力,同时增强了对行人目标特征的提取。为了更好地识别目标的边缘和重叠情况,提高回归精度,在损失函数中用SIoU代替CIoU,提升了遮挡情况下的检测率。根据在WiderPerson数据集上的实验,在保证检测速度的前提下,对比YOLOv7⁃tiny检测算法,平均检测精度提升了2个百分点。实验结果表明,经过改进的算法可以显著提高检测性能。 In high⁃density traffic scenarios for pedestrian detection,detection algorithms usually miss obscured and distant fuzzy pedestrians,while failing to balance detection accuracy and speed.To address these problems,an improved outdoor pedes⁃trian detection algorithm based on YOLOv7⁃tiny is proposed.The algorithm introduces the SENet attention mechanism,which sup⁃presses irrelevant information as a way to improve the ability of the feature map to express information,and at the same time en⁃hances the extraction of pedestrian target features.In order to better recognize the edges and overlapping of targets and improve the regression accuracy,SIoU is used instead of CIoU to improve the detection rate in the case of occlusion.According to the experi⁃ments on the WiderPerson dataset,the average detection accuracy is improved by 2 percentage comparing with the YOLOv7⁃tiny detection algorithm under the premise of guaranteeing the detection speed.The experimental results show that the improved algo⁃rithm can significantly improve the detection performance.
作者 潘兴好 郭彩萍 Pan Xinghao;Guo Caiping(College of Computer Science and Technology,Taiyuan Normal University,Jinzhong 030619,China;Department of Electronic Engineering,Taiyuan Institute of Technology,Taiyuan 030008,China)
出处 《现代计算机》 2024年第3期54-60,共7页 Modern Computer
关键词 行人检测 YOLOv7⁃tiny SENet SIoU pedestrian detection YOLOv7⁃tiny SENet SIoU
  • 相关文献

参考文献1

二级参考文献59

  • 1Lowe D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision, 2004, 60 (2) 91 110.
  • 2Dalai N, Triggs B. Histograms of oriented gradients for human detection[C]//Computer Vision and Pattern Recognition (CVPR), IEEE Computer Society Conference on. San Diego, USA: IEEE, 2005, 1 886-893.
  • 3Hinton G E, Salakhutdinov R R. Reducing the dimensionality of data with neural networks[J]. Science, 2006, 313(5786) : 504-507.
  • 4Hubel D H, Wiesel T N. Receptive fields, binocular interaction and functional architecture in the catrs visual cortex[J]. The Journal of Physiology, 1962, 160(1): 106-154.
  • 5Fukushima K, Miyake S. Neocognitron: A new algorithm for pattern recognition tolerant of deformations and shifts in posi- tion[J]. Pattern Recognition, 1982, 15(6): 455-469.
  • 6Ruck D W, Rogers S K, Kabrisky M. Feature selection using a multilayer perceptron[J]. Journal of Neural Network Com- puting, 1990, 2(2): 40-48.
  • 7Rumelhart D E, Hinton G E, Williams R J. Learning representations by back-propagating errors[J]. Nature, 1986,3231 533 538.
  • 8LeCun Y, Denker J S, Henderson D, et al. Handwritten digit recognition with a back-propagation network[C]//Advances in Neural Information Processing Systems. Colorado, USA Is. n. ], 1990: 396-404.
  • 9LeCun Y, Cortes C. MNIST handwritten digit database[EB/OL], http//yann, lecun, com/exdb/mnist, 2010.
  • 10Waibe[ A, Hanazawa T, Hinton G, et al. Phoneme recognition using time-delay neural networks[J]. Acoustics, Speech and Signal Processing, IEEF. Transactions on, 1989, 37(3): 328-339.

共引文献550

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部