TY - JOUR
T1 - TOD-CNN
T2 - An effective convolutional neural network for tiny object detection in sperm videos
AU - Zou, Shuojia
AU - Li, Chen
AU - Sun, Hongzan
AU - Xu, Peng
AU - Zhang, Jiawei
AU - Ma, Pingli
AU - Yao, Yudong
AU - Huang, Xinyu
AU - Grzegorzek, Marcin
N1 - Publisher Copyright:
© 2022 Elsevier Ltd
PY - 2022/7
Y1 - 2022/7
N2 - The detection of tiny objects in microscopic videos is a problematic point, especially in large-scale experiments. For tiny objects (such as sperms) in microscopic videos, current detection methods face challenges in fuzzy, irregular, and precise positioning of objects. In contrast, we present a convolutional neural network for tiny object detection (TOD-CNN) with an underlying data set of high-quality sperm microscopic videos (111 videos, > 278,000 annotated objects), and a graphical user interface (GUI) is designed to employ and test the proposed model effectively. TOD-CNN is highly accurate, achieving 85.60% AP50 in the task of real-time sperm detection in microscopic videos. To demonstrate the importance of sperm detection technology in sperm quality analysis, we carry out relevant sperm quality evaluation metrics and compare them with the diagnosis results from medical doctors.
AB - The detection of tiny objects in microscopic videos is a problematic point, especially in large-scale experiments. For tiny objects (such as sperms) in microscopic videos, current detection methods face challenges in fuzzy, irregular, and precise positioning of objects. In contrast, we present a convolutional neural network for tiny object detection (TOD-CNN) with an underlying data set of high-quality sperm microscopic videos (111 videos, > 278,000 annotated objects), and a graphical user interface (GUI) is designed to employ and test the proposed model effectively. TOD-CNN is highly accurate, achieving 85.60% AP50 in the task of real-time sperm detection in microscopic videos. To demonstrate the importance of sperm detection technology in sperm quality analysis, we carry out relevant sperm quality evaluation metrics and compare them with the diagnosis results from medical doctors.
KW - Convolutional neural network
KW - Image analysis
KW - Object detection
KW - Sperm microscopy video
UR - http://www.scopus.com/inward/record.url?scp=85129823935&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85129823935&partnerID=8YFLogxK
U2 - 10.1016/j.compbiomed.2022.105543
DO - 10.1016/j.compbiomed.2022.105543
M3 - Article
C2 - 35483229
AN - SCOPUS:85129823935
SN - 0010-4825
VL - 146
JO - Computers in Biology and Medicine
JF - Computers in Biology and Medicine
M1 - 105543
ER -