Related Experiment Video
Updated: Jan 18, 2026
13:19
Deep Neural Networks for Image-Based Dietary Assessment
Published on: March 13, 2021
9.9K
Dairy DigiD: a keypoint-based deep learning system for classifying dairy cattle by physiological and reproductive
Shubhangi Mahato1, Hanqing Bi2, Suresh Neethirajan1,3
1Faculty of Computer Science, Dalhousie University, Halifax, NS, Canada.
Frontiers in Artificial Intelligence
|September 8, 2025
Summary
Dairy DigiD uses AI facial recognition to classify dairy cows into four groups. This non-invasive system offers accurate, ethical livestock monitoring, improving precision dairy management.
Area of Science:
- Agricultural Science
- Artificial Intelligence
- Computer Vision
Background:
- Precision livestock farming needs non-invasive monitoring systems for cattle.
- Traditional methods like ear tags can cause discomfort and yield inconsistent data.
- Existing systems struggle with real-world farm conditions, impacting animal welfare.
Purpose of the Study:
- To introduce Dairy DigiD, a deep learning framework for biometric classification of dairy cattle using facial images.
- To categorize cows into four physiological groups: young, mature milking, pregnant, and dry cows.
- To develop an accurate and ethical alternative to traditional livestock identification methods.
Main Methods:
- Utilized a deep learning framework combining DenseNet121 for global image context and Detectron2 for facial analysis.
- Employed Detectron2's instance segmentation and keypoint detection on 30 facial landmarks for robust localization.
- Implemented cross-validation and explainability techniques to ensure biologically salient features guided classification.
Main Results:
- Detectron2 demonstrated superior adaptability in uncontrolled farm environments, achieving 93-98% classification accuracy.
- The keypoint-driven approach proved resilient to occlusions, lighting variations, and background heterogeneity.
- Explainability confirmed that biologically relevant facial features drove the classification outcomes, enhancing model transparency.
Conclusions:
- Dairy DigiD offers a significant advancement in automated livestock monitoring through an animal-centric AI approach.
- The system provides an ethical, accurate, and practical alternative to conventional identification methods.
- This framework sets a precedent for data-driven decision-making in precision dairy management.
Related Concept Videos
01:31Classification of Systems-II
460
Continuous-time systems have continuous input and output signals, with time measured continuously. These systems are generally defined by differential or algebraic equations. For instance, in an RC circuit, the relationship between input and output voltage is expressed through a differential equation derived from Ohm's law and the capacitor relation,
460
01:26Classification of Systems-I
552
Linearity is a system property characterized by a direct input-output relationship, combining homogeneity and additivity.
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
552
01:29Aggregates Classification
972
Aggregate classification is generally based on its size, petrographic characteristics, weight, and source. Size classification ranges from coarse to fine aggregates, defined by the size of the particles. Coarse aggregates are particles that do not pass through ASTM sieve No. 4, and aggregates that pass through the sieve are fine aggregates.
Petrographic classification groups aggregates based on common mineralogical characteristics. Some of the common mineral groups found in aggregates are...
Petrographic classification groups aggregates based on common mineralogical characteristics. Some of the common mineral groups found in aggregates are...
972