FSVM: A Few-Shot Threat Detection Method for X-ray Security Images
In recent years, automatic detection of threats in X-ray baggage has become important in security inspection. However, the training of threat detectors often requires extensive, well-annotated images, which are hard to procure, especially for rare contraband items. In this paper, a few-shot SVM-cons...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-04-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/8/4069 |
| _version_ | 1797603475390988288 |
|---|---|
| author | Cheng Fang Jiayue Liu Ping Han Mingrui Chen Dayu Liao |
| author_facet | Cheng Fang Jiayue Liu Ping Han Mingrui Chen Dayu Liao |
| author_sort | Cheng Fang |
| collection | DOAJ |
| description | In recent years, automatic detection of threats in X-ray baggage has become important in security inspection. However, the training of threat detectors often requires extensive, well-annotated images, which are hard to procure, especially for rare contraband items. In this paper, a few-shot SVM-constraint threat detection model, named FSVM is proposed, which aims at detecting unseen contraband items with only a small number of labeled samples. Rather than simply finetuning the original model, FSVM embeds a derivable SVM layer to back-propagate the supervised decision information into the former layers. A combined loss function utilizing SVM loss is also created as the additional constraint. We have evaluated FSVM on the public security baggage dataset SIXray, performing experiments on 10-shot and 30-shot samples under three class divisions. Experimental results show that compared with four common few-shot detection models, FSVM has the highest performance and is more suitable for complex distributed datasets (e.g., X-ray parcels). |
| first_indexed | 2024-03-11T04:32:39Z |
| format | Article |
| id | doaj.art-0a6833dcfdbd42a3a060213bbd1913d8 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T04:32:39Z |
| publishDate | 2023-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-0a6833dcfdbd42a3a060213bbd1913d82023-11-17T21:18:37ZengMDPI AGSensors1424-82202023-04-01238406910.3390/s23084069FSVM: A Few-Shot Threat Detection Method for X-ray Security ImagesCheng Fang0Jiayue Liu1Ping Han2Mingrui Chen3Dayu Liao4Tianjin Key Lab for Advanced Signal Processing, Civil Aviation University of China, Tianjin 300000, ChinaTianjin Key Lab for Advanced Signal Processing, Civil Aviation University of China, Tianjin 300000, ChinaTianjin Key Lab for Advanced Signal Processing, Civil Aviation University of China, Tianjin 300000, ChinaTianjin Key Lab for Advanced Signal Processing, Civil Aviation University of China, Tianjin 300000, ChinaTianjin Key Lab for Advanced Signal Processing, Civil Aviation University of China, Tianjin 300000, ChinaIn recent years, automatic detection of threats in X-ray baggage has become important in security inspection. However, the training of threat detectors often requires extensive, well-annotated images, which are hard to procure, especially for rare contraband items. In this paper, a few-shot SVM-constraint threat detection model, named FSVM is proposed, which aims at detecting unseen contraband items with only a small number of labeled samples. Rather than simply finetuning the original model, FSVM embeds a derivable SVM layer to back-propagate the supervised decision information into the former layers. A combined loss function utilizing SVM loss is also created as the additional constraint. We have evaluated FSVM on the public security baggage dataset SIXray, performing experiments on 10-shot and 30-shot samples under three class divisions. Experimental results show that compared with four common few-shot detection models, FSVM has the highest performance and is more suitable for complex distributed datasets (e.g., X-ray parcels).https://www.mdpi.com/1424-8220/23/8/4069X-ray imagesbaggage threat detectionfew-shot learningsupport vector machine |
| spellingShingle | Cheng Fang Jiayue Liu Ping Han Mingrui Chen Dayu Liao FSVM: A Few-Shot Threat Detection Method for X-ray Security Images Sensors X-ray images baggage threat detection few-shot learning support vector machine |
| title | FSVM: A Few-Shot Threat Detection Method for X-ray Security Images |
| title_full | FSVM: A Few-Shot Threat Detection Method for X-ray Security Images |
| title_fullStr | FSVM: A Few-Shot Threat Detection Method for X-ray Security Images |
| title_full_unstemmed | FSVM: A Few-Shot Threat Detection Method for X-ray Security Images |
| title_short | FSVM: A Few-Shot Threat Detection Method for X-ray Security Images |
| title_sort | fsvm a few shot threat detection method for x ray security images |
| topic | X-ray images baggage threat detection few-shot learning support vector machine |
| url | https://www.mdpi.com/1424-8220/23/8/4069 |
| work_keys_str_mv | AT chengfang fsvmafewshotthreatdetectionmethodforxraysecurityimages AT jiayueliu fsvmafewshotthreatdetectionmethodforxraysecurityimages AT pinghan fsvmafewshotthreatdetectionmethodforxraysecurityimages AT mingruichen fsvmafewshotthreatdetectionmethodforxraysecurityimages AT dayuliao fsvmafewshotthreatdetectionmethodforxraysecurityimages |