by Lu, J, Yan, W-Q and Nguyen, M
Abstract:
The traditional pedestrian detection is mostly based on global features of digital images. Nowadays, with the increase of computational power and processing capacity, deep neural networks (DNNs) acquires a high possibility to detect any objects, which has effectively led to a new era of machine learning. In this paper, we investigated a pedestrian detection tool using deep learning based on the YOLOv3 model. After a number of experiments conducted, our YOLOv3 model had shown to achieve 80.20% of accuracy in pedestrian detection with the speed of approximate 15 fps using GPU acceleration. Our direct contributions are: (1) data augment and collection, (2) adjusting deep neural network structures, and (3) superior performance in evaluations for our proposed deep learning model.
Reference:
Human behaviour recognition using deep learning (Lu, J, Yan, W-Q and Nguyen, M), In 2018 15th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS), IEEE, 2018.
Bibtex Entry:
@inproceedings{lu2018humanlearning,
author = {Lu, J and Yan, W-Q and Nguyen, M},
booktitle = {2018 15th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS)},
organization = {Auckland},
publisher = {IEEE},
title = {Human behaviour recognition using deep learning},
url = {https://ieeexplore.ieee.org/document/8639413},
year = {2018},
abstract = {The traditional pedestrian detection is mostly based on global features of digital images. Nowadays, with the increase of computational power and processing capacity, deep neural networks (DNNs) acquires a high possibility to detect any objects, which has effectively led to a new era of machine learning. In this paper, we investigated a pedestrian detection tool using deep learning based on the YOLOv3 model. After a number of experiments conducted, our YOLOv3 model had shown to achieve 80.20% of accuracy in pedestrian detection with the speed of approximate 15 fps using GPU acceleration. Our direct contributions are: (1) data augment and collection, (2) adjusting deep neural network structures, and (3) superior performance in evaluations for our proposed deep learning model.},
doi = {10.1109/AVSS.2018.8639413},
startyear = {2018},
startmonth = {Nov},
startday = {27},
finishyear = {2018},
finishmonth = {Nov},
finishday = {30},
isbn = {978-1-5386-9294-3},
conference = {The 4th International Workshop on Digital Crime and Forensics},
}