Deep Learning Approaches for Pedestrian Detection: A Review of Recent Advances
Abstract
including occlusion, multi-scale target variations, illumination fluctuations, and dense crowd scenarios. This paper systematically reviews
the three-stage evolutionary trajectory of pedestrian detection technologies, analyzes mainstream architectures exemplified by the YOLO and
DETR series, summarizes core optimization techniques, and compares algorithmic performance in terms of accuracy, speed, and adaptability.
It further discusses prevailing limitations and outlines prospective research directions, thereby providing a valuable reference for algorithm
optimization and engineering implementation.
Keywords
Full Text:
PDFReferences
[1] Ma Fan. Talking about the policy of extra subtraction of R & D costs of small and medium-sized businesses [J]. Finance, 2023, (15): 51-
53.
[2] M SUKKAR, R JADEJA, M SHUKLA, et al. Review of deep learning methods to detect pedestrians on self-driving systems[J]. IEEE
Access, 2025, 13: 3994-4007.
[3] LI C, LI L, JIANG H, et al. YOLOv6: One-stage object detection model of industrial use[J]. Pattern Recognition Letters, 2023, 172: 1-10.
[4] Z, LI, ZHANG Y, WANG C, et al. A better pedestrian detection algorithm using YOLOv5s[J]. Journal of Advanced Computational Intelligence and Intelligent Informatics, 2024, 28(4): 768-775.
[5] K CHEN, Z ZHANG, W Zhang and others. Pedestrian Detection based on Transformers with Increased Interaction of Features[J]. IEEE
Transactions on Intelligent Transportation Systems, 2024, 25: 1-12.
[6] LIANG S., WANG W., WANG J., and others. Pedestrian and vehicle detection using pruning YOLOv4 with cloud-edge collaboration[J].
Computer Modeling in Engineering and Sciences, 2023, 137(2): 2025-2047.
[7] Q LI, C ZHANG, Q HU, et al. Stabilization of multispectral pedestrian detection based on evidential hybrid fusion[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2024, 34(4): 3017-3029.
[8] NIE L, LU M, HE Z, et al. Pedestrian detection using multispectral techniques by means of complementation and enhancement of features [J]. IET Intelligent Transport Systems, 2024, 18 (11): 2166-2177.
[9] WANG L, BAI J, WANG P, et al. Pedestrian detection algorithm in industrial scene with improved YOLOv7-Tiny[J]. IEEJ Transactions
on Electrical and Electronic Engineering, 2024, 19(7): 1203-1215.
[10] W. HANIF M, LI Z, YU Z. A lightweight object detection system in the context of edge computing to be used by the mining industry[J].
IET Image Processing, 2024, 18(13): 4005-4022.
DOI: http://dx.doi.org/10.70711/aitr.v4i4.10159
Refbacks
- There are currently no refbacks.
