Sustainable Self-Training Pig Detection System with Augmented Single Labeled Target Data for Solving Domain Shift Problem

As global pork consumption rises, livestock farms increasingly adopt deep learning-based automated monitoring systems for efficient pigsty management. Typically, a system applies a pre-trained model on a source domain to a target domain. However, real pigsty environments differ significantly from ex...

Full description

Saved in:
Bibliographic Details
Main Authors: Junhee Lee, Heechan Chae, Seungwook Son, Jongwoong Seo, Yooil Suh, Jonguk Lee, Yongwha Chung, Daihee Park
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/11/3406
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850129114145488896
author Junhee Lee
Heechan Chae
Seungwook Son
Jongwoong Seo
Yooil Suh
Jonguk Lee
Yongwha Chung
Daihee Park
author_facet Junhee Lee
Heechan Chae
Seungwook Son
Jongwoong Seo
Yooil Suh
Jonguk Lee
Yongwha Chung
Daihee Park
author_sort Junhee Lee
collection DOAJ
description As global pork consumption rises, livestock farms increasingly adopt deep learning-based automated monitoring systems for efficient pigsty management. Typically, a system applies a pre-trained model on a source domain to a target domain. However, real pigsty environments differ significantly from existing public datasets regarding lighting conditions, camera angles, and animal density. These discrepancies result in a substantial domain shift, leading to severe performance degradation. Additionally, due to variations in the structure of pigsties, pig breeds, and sizes across farms, it is practically challenging to develop a single generalized model that can be applied to all environments. Overcoming this limitation through large-scale labeling presents considerable burdens in terms of time and cost. To address the degradation issue, this study proposes a self-training-based domain adaptation method that utilizes a single label on target (SLOT) sample from the target domain, a genetic algorithm (GA)-based data augmentation search (DAS) designed explicitly for SLOT data to optimize the augmentation parameters, and a super-low-threshold strategy to include low-confidence-scored pseudo-labels during self-training. The proposed system consists of the following three modules: (1) data collection module; (2) preprocessing module that selects key frames and extracts SLOT data; and (3) domain-adaptive pig detection module that applies DAS to SLOT data to generate optimized augmented data, which are used to train the base model. Then, the trained base model is improved through self-training, where a super-low threshold is applied to filter pseudo-labels. The experimental results show that the proposed system significantly improved the average precision (AP) from 36.86 to 90.62 under domain shift conditions, which achieved a performance close to fully supervised learning while relying solely on SLOT data. The proposed system maintained a robust detection performance across various pig-farming environments and demonstrated stable performance under domain shift conditions, validating its feasibility for real-world applications.
format Article
id doaj-art-e3b3d8a6436f4fe8837f04ada6d606c3
institution OA Journals
issn 1424-8220
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-e3b3d8a6436f4fe8837f04ada6d606c32025-08-20T02:33:06ZengMDPI AGSensors1424-82202025-05-012511340610.3390/s25113406Sustainable Self-Training Pig Detection System with Augmented Single Labeled Target Data for Solving Domain Shift ProblemJunhee Lee0Heechan Chae1Seungwook Son2Jongwoong Seo3Yooil Suh4Jonguk Lee5Yongwha Chung6Daihee Park7Info Valley Korea Co., Ltd., Anyang 14067, Republic of KoreaInfo Valley Korea Co., Ltd., Anyang 14067, Republic of KoreaInfo Valley Korea Co., Ltd., Anyang 14067, Republic of KoreaDepartment of Computer Convergence Software, Korea University, Sejong 30019, Republic of KoreaInfo Valley Korea Co., Ltd., Anyang 14067, Republic of KoreaDepartment of Computer Convergence Software, Korea University, Sejong 30019, Republic of KoreaDepartment of Computer Convergence Software, Korea University, Sejong 30019, Republic of KoreaDepartment of Computer Convergence Software, Korea University, Sejong 30019, Republic of KoreaAs global pork consumption rises, livestock farms increasingly adopt deep learning-based automated monitoring systems for efficient pigsty management. Typically, a system applies a pre-trained model on a source domain to a target domain. However, real pigsty environments differ significantly from existing public datasets regarding lighting conditions, camera angles, and animal density. These discrepancies result in a substantial domain shift, leading to severe performance degradation. Additionally, due to variations in the structure of pigsties, pig breeds, and sizes across farms, it is practically challenging to develop a single generalized model that can be applied to all environments. Overcoming this limitation through large-scale labeling presents considerable burdens in terms of time and cost. To address the degradation issue, this study proposes a self-training-based domain adaptation method that utilizes a single label on target (SLOT) sample from the target domain, a genetic algorithm (GA)-based data augmentation search (DAS) designed explicitly for SLOT data to optimize the augmentation parameters, and a super-low-threshold strategy to include low-confidence-scored pseudo-labels during self-training. The proposed system consists of the following three modules: (1) data collection module; (2) preprocessing module that selects key frames and extracts SLOT data; and (3) domain-adaptive pig detection module that applies DAS to SLOT data to generate optimized augmented data, which are used to train the base model. Then, the trained base model is improved through self-training, where a super-low threshold is applied to filter pseudo-labels. The experimental results show that the proposed system significantly improved the average precision (AP) from 36.86 to 90.62 under domain shift conditions, which achieved a performance close to fully supervised learning while relying solely on SLOT data. The proposed system maintained a robust detection performance across various pig-farming environments and demonstrated stable performance under domain shift conditions, validating its feasibility for real-world applications.https://www.mdpi.com/1424-8220/25/11/3406monitoring systemobject detectiondomain shiftdata augmentationgenetic algorithmself-training
spellingShingle Junhee Lee
Heechan Chae
Seungwook Son
Jongwoong Seo
Yooil Suh
Jonguk Lee
Yongwha Chung
Daihee Park
Sustainable Self-Training Pig Detection System with Augmented Single Labeled Target Data for Solving Domain Shift Problem
Sensors
monitoring system
object detection
domain shift
data augmentation
genetic algorithm
self-training
title Sustainable Self-Training Pig Detection System with Augmented Single Labeled Target Data for Solving Domain Shift Problem
title_full Sustainable Self-Training Pig Detection System with Augmented Single Labeled Target Data for Solving Domain Shift Problem
title_fullStr Sustainable Self-Training Pig Detection System with Augmented Single Labeled Target Data for Solving Domain Shift Problem
title_full_unstemmed Sustainable Self-Training Pig Detection System with Augmented Single Labeled Target Data for Solving Domain Shift Problem
title_short Sustainable Self-Training Pig Detection System with Augmented Single Labeled Target Data for Solving Domain Shift Problem
title_sort sustainable self training pig detection system with augmented single labeled target data for solving domain shift problem
topic monitoring system
object detection
domain shift
data augmentation
genetic algorithm
self-training
url https://www.mdpi.com/1424-8220/25/11/3406
work_keys_str_mv AT junheelee sustainableselftrainingpigdetectionsystemwithaugmentedsinglelabeledtargetdataforsolvingdomainshiftproblem
AT heechanchae sustainableselftrainingpigdetectionsystemwithaugmentedsinglelabeledtargetdataforsolvingdomainshiftproblem
AT seungwookson sustainableselftrainingpigdetectionsystemwithaugmentedsinglelabeledtargetdataforsolvingdomainshiftproblem
AT jongwoongseo sustainableselftrainingpigdetectionsystemwithaugmentedsinglelabeledtargetdataforsolvingdomainshiftproblem
AT yooilsuh sustainableselftrainingpigdetectionsystemwithaugmentedsinglelabeledtargetdataforsolvingdomainshiftproblem
AT jonguklee sustainableselftrainingpigdetectionsystemwithaugmentedsinglelabeledtargetdataforsolvingdomainshiftproblem
AT yongwhachung sustainableselftrainingpigdetectionsystemwithaugmentedsinglelabeledtargetdataforsolvingdomainshiftproblem
AT daiheepark sustainableselftrainingpigdetectionsystemwithaugmentedsinglelabeledtargetdataforsolvingdomainshiftproblem