Minseo Jo

Minseo Jo

orcid

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 November 2025 in IEEE Access
Verified
Journal Article
Open access
MJMinseo Jo
Minseo Jo
SSSookeun Song
Sookeun Song
HYHyunbean Yi
Hyunbean Yi
This study proposes a method for estrus prediction in sows using object detection and convolutional recurrent neural network (CRNN) models. Previous estrus prediction models have exhibited inconsistent performance due to variations in posture-transition event frequency and ambiguity in detecting transitions. We observed that when a sow changes its posture once, multiple detections of posture changes often occur, leading to an abnormally high frequency of events. To address this issue, the time-series data were resampled at uniform intervals. The resampled time-series data were processed using a double moving average, which served as the input for the prediction model. In 142 of 175 sows, the activity level during the 24 h preceding the estrus detection point by farm managers showed an increasing trend compared with the previous time period. This observed activity trend is consistent with the behavioral characteristics of sows during estrus reported in previous studies. Estrus prediction using the newly constructed time-series data and CRNN model achieved an accuracy of approximately 86%. The proposed method is expected to facilitate more efficient estrus prediction in pig farms, thereby reducing the associated costs and labor.

10 September 2025 in Agriculture
Verified
Journal Article
Open access
PDF file
SSSookeun Song
Sookeun Song
MJMinseo Jo
Minseo Jo
BLBong Kuk Lee
Bong Kuk Lee
SLSangkeum Lee
Sangkeum Lee
HYHyunbean Yi
Hyunbean Yi
Determining sow (Sus scrofa domesticus Erxleben) estrus status requires considerable labor investment, and continuous real-time monitoring is impractical. Workers typically identify estrus at scheduled intervals and determine artificial insemination timing based on experience. However, experience-based methods are subjective, vary with operator expertise, and impede standardized management in large-scale farms. This study employs cameras and deep learning to detect sows and analyze postural changes, enabling estrus detection and optimal insemination timing prediction. Experimental results indicate that the proposed method achieved an accuracy of 70% (42/60), where the recommended insemination timing differed by less than 24 h from human decisions. This approach facilitates data-driven estrus detection and insemination scheduling, potentially reducing labor intensity and improving reproductive outcomes, particularly beneficial for labor-intensive and large-scale swine production systems.