TOD-CNN: An effective convolutional neural network for tiny object detection in sperm videos

Shuojia Zou, Chen Li, Hongzan Sun, Peng Xu, Jiawei Zhang, Pingli Ma, Yudong Yao, Xinyu Huang, Marcin Grzegorzek

Research output: Contribution to journalArticlepeer-review

26 Scopus citations

Abstract

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.

Original languageEnglish
Article number105543
JournalComputers in Biology and Medicine
Volume146
DOIs
StatePublished - Jul 2022

Keywords

  • Convolutional neural network
  • Image analysis
  • Object detection
  • Sperm microscopy video

Fingerprint

Dive into the research topics of 'TOD-CNN: An effective convolutional neural network for tiny object detection in sperm videos'. Together they form a unique fingerprint.

Cite this