Related Experiment Video
Updated: Jun 2, 2025
High Throughput Assay to Examine Egg-Laying Preferences of Individual Drosophila melanogaster
Published on: March 24, 2016
JTF-SqueezeNet: A SqueezeNet network based on joint time-frequency data representation for egg-laying detection in
Siting Lv1, Yuanyang Mao1, Youfu Liu1
1College of Mathematics Informatics, South China Agricultural University, Guangzhou 510642, China; Key Laboratory of Smart Agricultural Technology in Tropical South China, Ministry of Agriculture and Rural Affairs, Guangzhou 510642, China; Guangdong Engineering Research Center of Agricultural Big Data, Guangzhou 510642, China.
Wearable sensors and joint time-frequency representation (TFR) enable accurate duck egg-laying detection. This method improves breeder duck efficiency by identifying egg-laying status with high accuracy and minimal inference time.
Area of Science:
- Animal Science
- Biotechnology
- Machine Learning
Background:
- Accurate detection of egg-laying in breeder ducks is vital for optimizing production efficiency.
- Current methods are often costly and require controlled environments, limiting practical application.
Purpose of the Study:
- To develop a cost-effective and accurate method for individual duck egg-laying detection using wearable sensors.
- To leverage joint time-frequency representation (TFR) for analyzing sensor data and identifying egg-laying events.
Main Methods:
- Continuous monitoring of duck activity via X-axis acceleration data from wearable sensors.
- Application of Short-Time Fourier Transform (STFT) to generate spectrograms from sensor data.
- Utilizing SqueezeNet for feature detection in spectrograms and Kalman filtering for status prediction.
Main Results:
- Achieved 95.73% accuracy in detecting egg-laying status with a 2.15 ms inference time per window.
- Identified egg-laying periods with 92.19% accuracy, 93.57% precision, and 91.95% recall.
- Demonstrated the scalability of TFR for reducing computational complexity.
Conclusions:
- The proposed sensor-based TFR method offers a highly accurate and efficient solution for monitoring duck egg-laying.
- This approach has the potential to significantly improve breeder duck management and productivity.
- Further exploration of TFR scalability can enhance computational efficiency for real-world deployment.
