The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Analisis imej berbutir halus, seperti pendekatan tahap piksel, meningkatkan pengesanan ancaman dalam imej keselamatan sinar-x. Dalam tetapan praktikal, kos untuk mendapatkan anotasi tahap piksel lengkap meningkat dengan ketara, yang boleh dikurangkan dengan melabelkan separa set data. Walau bagaimanapun, pengendalian set data berlabel separa boleh membawa kepada latihan rangkaian berbilang peringkat yang rumit. Dalam makalah ini, kami mencadangkan rangka kerja pemisahan objek hujung ke hujung baharu yang melatih rangkaian tunggal pada set data berlabel separa sambil turut mengurangkan ketidakseimbangan kelas yang wujud pada peringkat data dan cadangan objek. Keputusan empirikal menunjukkan peningkatan yang ketara berbanding pendekatan sedia ada.
Joanna Kazzandra DUMAGPI
Kwangwoon University
Yong-Jin JEONG
Kwangwoon University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Salinan
Joanna Kazzandra DUMAGPI, Yong-Jin JEONG, "End-to-End Object Separation for Threat Detection in Large-Scale X-Ray Security Images" in IEICE TRANSACTIONS on Information,
vol. E105-D, no. 10, pp. 1807-1811, October 2022, doi: 10.1587/transinf.2022EDL8019.
Abstract: Fine-grained image analysis, such as pixel-level approaches, improves threat detection in x-ray security images. In the practical setting, the cost of obtaining complete pixel-level annotations increases significantly, which can be reduced by partially labeling the dataset. However, handling partially labeled datasets can lead to training complicated multi-stage networks. In this paper, we propose a new end-to-end object separation framework that trains a single network on a partially labeled dataset while also alleviating the inherent class imbalance at the data and object proposal level. Empirical results demonstrate significant improvement over existing approaches.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2022EDL8019/_p
Salinan
@ARTICLE{e105-d_10_1807,
author={Joanna Kazzandra DUMAGPI, Yong-Jin JEONG, },
journal={IEICE TRANSACTIONS on Information},
title={End-to-End Object Separation for Threat Detection in Large-Scale X-Ray Security Images},
year={2022},
volume={E105-D},
number={10},
pages={1807-1811},
abstract={Fine-grained image analysis, such as pixel-level approaches, improves threat detection in x-ray security images. In the practical setting, the cost of obtaining complete pixel-level annotations increases significantly, which can be reduced by partially labeling the dataset. However, handling partially labeled datasets can lead to training complicated multi-stage networks. In this paper, we propose a new end-to-end object separation framework that trains a single network on a partially labeled dataset while also alleviating the inherent class imbalance at the data and object proposal level. Empirical results demonstrate significant improvement over existing approaches.},
keywords={},
doi={10.1587/transinf.2022EDL8019},
ISSN={1745-1361},
month={October},}
Salinan
TY - JOUR
TI - End-to-End Object Separation for Threat Detection in Large-Scale X-Ray Security Images
T2 - IEICE TRANSACTIONS on Information
SP - 1807
EP - 1811
AU - Joanna Kazzandra DUMAGPI
AU - Yong-Jin JEONG
PY - 2022
DO - 10.1587/transinf.2022EDL8019
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E105-D
IS - 10
JA - IEICE TRANSACTIONS on Information
Y1 - October 2022
AB - Fine-grained image analysis, such as pixel-level approaches, improves threat detection in x-ray security images. In the practical setting, the cost of obtaining complete pixel-level annotations increases significantly, which can be reduced by partially labeling the dataset. However, handling partially labeled datasets can lead to training complicated multi-stage networks. In this paper, we propose a new end-to-end object separation framework that trains a single network on a partially labeled dataset while also alleviating the inherent class imbalance at the data and object proposal level. Empirical results demonstrate significant improvement over existing approaches.
ER -