Artificial Intelligence Applications for Citizen Science Cricket Cell Counting
University
Shawnee State University
Major
Computer Science
Student Type
Undergraduate Student
Presentation Types
Oral Presentation (Live)
Keywords:
citizen science, biology, artificial intelligence, computer vision
Abstract
Studies have shown that crickets living in higher levels of noise pollution may have a lower chance of reproducing, causing devastating effects on the cricket population. Recent work involving crickets in the Hawaiian Islands has studied this by collecting sperm samples to compare the ratio of living to dead cells under the hypothesis that crickets in louder environments could be prone to infertility. Currently, the cricket study relies on crowd sourced citizen scientist help to count the number of cells per sample by hand and record their results on their website. Problematically, citizen science volunteers may not be qualified to recognize cells in a sample. This project aims to address this research need through Artificial Intelligence (AI) to automate the process of counting these cells. With the implementation of our AI, we have created a much faster and more reliable method for completing this tedious task. Analysis is conducted to compare human-annotated cricket cell counts with AI-generated annotations, assessing both consistency and the potential for detecting previously uncounted cells. The result of this project is a repeatable and reliable system which does not require possibly untrained crowd sourced solutions that speeds up a key, but tedious task, in biology and helps shorten the path to scientific discovery.
Human and Animal Subjects
no
IRB or IACUC Approval
no
Faculty Mentor Name
Dr. Trevor Bihl
Faculty Mentor Title
Adjunct Faculty
Faculty Mentor Department
Engineering Technologies
Recommended Citation
Knisley, Hannah and Bihl, Trevor, "Artificial Intelligence Applications for Citizen Science Cricket Cell Counting" (2025). Celebration of Scholarship. 2.
https://digitalcommons.shawnee.edu/cos/2025/session1/2
Location
LIB 204
Artificial Intelligence Applications for Citizen Science Cricket Cell Counting
LIB 204
Studies have shown that crickets living in higher levels of noise pollution may have a lower chance of reproducing, causing devastating effects on the cricket population. Recent work involving crickets in the Hawaiian Islands has studied this by collecting sperm samples to compare the ratio of living to dead cells under the hypothesis that crickets in louder environments could be prone to infertility. Currently, the cricket study relies on crowd sourced citizen scientist help to count the number of cells per sample by hand and record their results on their website. Problematically, citizen science volunteers may not be qualified to recognize cells in a sample. This project aims to address this research need through Artificial Intelligence (AI) to automate the process of counting these cells. With the implementation of our AI, we have created a much faster and more reliable method for completing this tedious task. Analysis is conducted to compare human-annotated cricket cell counts with AI-generated annotations, assessing both consistency and the potential for detecting previously uncounted cells. The result of this project is a repeatable and reliable system which does not require possibly untrained crowd sourced solutions that speeds up a key, but tedious task, in biology and helps shorten the path to scientific discovery.
