TY - JOUR
T1 - A Survey of Crime Scene Investigation Image Retrieval Using Deep Learning
AU - Liu, Ying
AU - Zhou, Aodong
AU - Xue, Jize
AU - Xu, Zhijie
N1 - Publisher Copyright:
© 2024 Beijing Institute of Technology. All rights reserved.
PY - 2024/8
Y1 - 2024/8
N2 - Crime scene investigation (CSI) image is key evidence carrier during criminal investigation, in which CSI image retrieval can assist the public police to obtain criminal clues. Moreover, with the rapid development of deep learning, data-driven paradigm has become the mainstream method of CSI image feature extraction and representation, and in this process, datasets provide effective support for CSI retrieval performance. However, there is a lack of systematic research on CSI image retrieval methods and datasets. Therefore, we present an overview of the existing works about one-class and multi-class CSI image retrieval based on deep learning. According to the research, based on their technical functionalities and implementation methods, CSI image retrieval is roughly classified into five categories: feature representation, metric learning, generative adversarial networks, autoencoder networks and attention networks. Furthermore, We analyzed the remaining challenges and discussed future work directions in this field.
AB - Crime scene investigation (CSI) image is key evidence carrier during criminal investigation, in which CSI image retrieval can assist the public police to obtain criminal clues. Moreover, with the rapid development of deep learning, data-driven paradigm has become the mainstream method of CSI image feature extraction and representation, and in this process, datasets provide effective support for CSI retrieval performance. However, there is a lack of systematic research on CSI image retrieval methods and datasets. Therefore, we present an overview of the existing works about one-class and multi-class CSI image retrieval based on deep learning. According to the research, based on their technical functionalities and implementation methods, CSI image retrieval is roughly classified into five categories: feature representation, metric learning, generative adversarial networks, autoencoder networks and attention networks. Furthermore, We analyzed the remaining challenges and discussed future work directions in this field.
KW - crime scene investigation (CSI) image
KW - deep learning
KW - image retrieval
UR - http://www.scopus.com/inward/record.url?scp=85207041932&partnerID=8YFLogxK
U2 - 10.15918/j.jbit1004-0579.2023.152
DO - 10.15918/j.jbit1004-0579.2023.152
M3 - Article
AN - SCOPUS:85207041932
SN - 1004-0579
VL - 33
SP - 271
EP - 286
JO - Journal of Beijing Institute of Technology (English Edition)
JF - Journal of Beijing Institute of Technology (English Edition)
IS - 4
ER -