Study on Pedestrian Detection Based on an Improved YOLOv4 Algorithm

Pedestrian detection, which is widely used in automatic driving and pedestrian analysis, has always been a hot research topic in the fields of artificial intelligence and computer vision. With the development of deep learning, pedestrian detectors are becoming more accurate and faster. However, most...

Full description

Saved in:
Bibliographic Details
Published in2020 IEEE 6th International Conference on Computer and Communications (ICCC) pp. 1198 - 1202
Main Authors Boyuan, Wen, Muqing, Wu
Format Conference Proceeding
LanguageEnglish
Published IEEE 11.12.2020
Subjects
activation function
Clustering algorithms
Computational modeling
Deep learning
Detectors
Feature extraction
K-means clustering
Neck
pedestrian detection
Real-time systems
spatial pyramid pooling
YOLOv4 algorithm
Online AccessGet full text
DOI10.1109/ICCC51575.2020.9344983

Cover

More Information
Summary:Pedestrian detection, which is widely used in automatic driving and pedestrian analysis, has always been a hot research topic in the fields of artificial intelligence and computer vision. With the development of deep learning, pedestrian detectors are becoming more accurate and faster. However, most of them can't strike a balance between accuracy and speed well. Therefore, in this study, we propose a pedestrian detection model based on an improved YOLOv4 algorithm which concerns both detection accuracy and efficiency. The detection model combines a new type of SPP (Spatial Pyramid Pooling) network and K-means clustering algorithm with YOLOv4 model for easier feature extraction. Furthermore, Mish activation function is applied in the neck of the detection model, replacing Leaky ReLU activation function to improve the detection performance. Our pedestrian detector achieves excellent results on the Caltech pedestrian dataset: 84.7% AP at a real-time speed of 36.4 FPS on Titan XP.
DOI:10.1109/ICCC51575.2020.9344983