Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

StomaYOLO: A Lightweight Maize Phenotypic Stomatal Cell Detector Based on Multi-Task Training.

Plants (Basel, Switzerland)·2025
Same author

YOLOv8A-SD: A Segmentation-Detection Algorithm for Overlooking Scenes in Pig Farms.

Animals : an open access journal from MDPI·2025
Same author

The Rapid Non-Destructive Differentiation of Different Varieties of Rice by Fluorescence Hyperspectral Technology Combined with Machine Learning.

Molecules (Basel, Switzerland)·2024
Same author

[Case-control study on open reduction and internal fixation for the treatment of humeral surgical neck fracture in aged patients].

Zhongguo gu shang = China journal of orthopaedics and traumatology·2015
Same author

Insights into the interaction between Cren7 and DNA: the role of loop β3-β4.

Extremophiles : life under extreme conditions·2015
Same author

Metatranscriptomic analyses of plant cell wall polysaccharide degradation by microorganisms in the cow rumen.

Applied and environmental microbiology·2014

Related Experiment Video

Updated: Sep 3, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

4.0K

Efficient Detection Method of Pig-Posture Behavior Based on Multiple Attention Mechanism.

Li Huang1, Lijia Xu1, Yuchao Wang1

  • 1College of Mechanical and Electrical Engineering, Sichuan Agriculture University, Ya'an 625014, China.

Computational Intelligence and Neuroscience
|July 26, 2022
PubMed
Summary

This study introduces the HE-Yolo model for accurate real-time pig posture and behavior detection in complex environments. The novel model demonstrates superior precision and robustness compared to existing methods, enhancing animal welfare monitoring.

More Related Videos

Noninvasive, In-pen Approach Test for Laboratory-housed Pigs
06:30

Noninvasive, In-pen Approach Test for Laboratory-housed Pigs

Published on: June 5, 2019

8.5K
A Method for Investigating Change Blindness in Pigeons Columba Livia
06:14

A Method for Investigating Change Blindness in Pigeons Columba Livia

Published on: September 7, 2018

6.5K

Related Experiment Videos

Last Updated: Sep 3, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

4.0K
Noninvasive, In-pen Approach Test for Laboratory-housed Pigs
06:30

Noninvasive, In-pen Approach Test for Laboratory-housed Pigs

Published on: June 5, 2019

8.5K
A Method for Investigating Change Blindness in Pigeons Columba Livia
06:14

A Method for Investigating Change Blindness in Pigeons Columba Livia

Published on: September 7, 2018

6.5K

Area of Science:

  • Computer Vision
  • Animal Science
  • Machine Learning

Background:

  • Traditional pig posture and behavior detection methods lack precision and robustness in complex farming environments.
  • Real-time monitoring of enclosure pig behaviors is crucial for health assessment and welfare management.

Purpose of the Study:

  • To design and validate the HE-Yolo (High-effect Yolo) model for precise and robust real-time recognition of enclosure pig posture behaviors.
  • To improve upon existing detection models by integrating advanced feature extraction and attention mechanisms.

Main Methods:

  • Developed the HE-Yolo model by enhancing the Darknet-53 feature extraction network with Depthwise separable convolution (DSC), h-switch activation, and a Contrary residual structure (C-Res) unit.
  • Integrated a Dual Attention Mechanism (DAM) combining channel and spatial attention, further incorporated into a Dual Attention Block (DAB).
  • Utilized K-means clustering for anchor frame optimization and trained the model on a dataset of 2912 pig images, evaluating performance using precision (P), recall (R), AP, and mAP.

Main Results:

  • The HE-Yolo model achieved high AP values for recognizing standing (99.25%), sitting (98.41%), prone (94.43%), and sidling (97.63%) pig postures.
  • Compared to Yolo v3, SSD, and Faster R-CNN, HE-Yolo demonstrated increased mAP values by 5.61%, 4.65%, and 0.57%, respectively, with a fast recognition time of 0.045s per frame.
  • HE-Yolo showed improved mAP values (4.04%, 4.94%, 1.76%) in challenging conditions like foreign body occlusion and pig adhesion, and under varying lighting.

Conclusions:

  • The HE-Yolo model offers high precision and robustness for recognizing pig posture behaviors, outperforming existing methods.
  • The model exhibits strong generalization capabilities and luminance robustness, making it suitable for complex farming environments.
  • HE-Yolo provides effective technical support for real-time monitoring of enclosure pig behaviors and physiological health.