Figure - available from: Animals
This content is subject to copyright.
Source publication
Simple Summary
Traditionally, farmers are unable to pay enough attention to individual livestock. An increasing number of sensors are being used to monitor animal behavior, early disease detection, and evaluation of animal welfare. In this study, we used machine learning algorithms to identify multiple unitary behaviors and movements of dairy cattl...
Similar publications
Diabetes is a disease attacking the endocrine system characterized by high blood sugar levels. International Diabetes Federation (IDF) estimates that there were 451 million people with diabetes globally in 2017. Without treatment, this number is expected to rise to 693 million by 2045. One method for preventing increases in the number of diabetics...
Citations
... With the continuous development of modern agriculture and animal husbandry, the increasing attention to sustainable agriculture and animal welfare has made dairy cow health management increasingly important [1,2]. The identification of cow behavior has become critical as cow behavioral patterns are closely related to cow health status, as well as to the yield and quality of dairy products [3][4][5]. ...
The behavioral changes of dairy cows directly reflect their health status, and observing the behavioral changes of dairy cows can provide a scientific basis for dairy farms so managers can take timely measures to intervene and effectively prevent diseases. Because of the complex background, multi-scale behavior changes of dairy cows, similar behavior, and difficulty in detecting small targets in the actual dairy farm environment, this study proposes a dairy cow behavior recognition algorithm, DMSF-YOLO, based on dynamic mechanism and multi-scale feature fusion, which can quickly and accurately identify the lying, standing, walking, eating, drinking and mounting behaviors of dairy cows. For the problem in multi-scale behavior changes of dairy cows, a multi-scale convolution module (MSFConv) is designed, and some C3k2 modules of the backbone network and neck network are replaced with MSFConv, which can extract cow behavior information of different scales and perform multi-scale feature fusion. Secondly, the C2BRA multi-scale feature extraction module is designed to replace the C2PSA module, which can dynamically select the important areas according to the two-layer routing attention mechanism to extract feature information at different scales and enhance the multi-scale feature extraction capability of the model, and the same time inhibit the interference of the background information to improve the small target detection capability of the model. Finally, the Dynamic Head detection head is introduced to enhance the model’s scale, spatial location, and perception of different tasks, enhance the capacity to extract similar behavioral features of cows, and improve the model’s performance in detecting cow multi-scale behaviors in complex environments. The proposed DMSF-YOLO algorithm is experimentally validated on a self-constructed cow behavior dataset, and the experimental results show that the DMSF-YOLO model improves the precision (P), recall (R), mAP50, and F1 values by 2.4%, 3%, 1.6%, and 2.7%, respectively, and the FPS value is also high. The model can suppress the interference of background information, dynamically extract multi-scale features, perform feature fusion, distinguish similar behaviors of cows, enhance the capacity to detect small targets, and significantly improve the recognition accuracy and overall performance of the model. This model can satisfy the need to quickly and accurately identify cow behavior in actual dairy farm environments.
... However, their use for specifically detecting JM events poses challenges due to the limited discriminatory power of the signals captured for this purpose . A variety of approaches have been explored, including the use of accelerometers (Tani et al., 2013;Oudshoorn et al., 2013;Bloch et al., 2023), accelerometers and gyroscopes (referred to as IMUs) (Andriamandroso et al., 2015;Li et al., 2022), and accelerometers, gyroscopes, and magnetometers (referred to as inertial and magnetic measurement units) . ...
Monitoring feeding behaviour is a relevant task for efficient herd management and the effective use of available resources in grazing cattle. The ability to automatically recognise animals' feeding activities through the identification of specific jaw movements allows for the improvement of diet formulation, as well as early detection of metabolic problems and symptoms of animal discomfort, among other benefits. The use of sensors to obtain signals for such monitoring has become popular in the last two decades. The most frequently employed sensors include accelerometers, microphones, and cameras, each with its own set of advantages and drawbacks. An unexplored aspect is the simultaneous use of multiple sensors with the aim of combining signals in order to enhance the precision of the estimations. In this direction, this work introduces a deep neural network based on the fusion of acoustic and inertial signals, composed of convolutional, recurrent, and dense layers. The main advantage of this model is the combination of signals through the automatic extraction of features independently from each of them. The model has emerged from an exploration and comparison of different neural network architectures proposed in this work, which carry out information fusion at different levels. Feature-level fusion has outperformed data and decision-level fusion by at least a 0.14 based on the F1-score metric. Moreover, a comparison with state-of-the-art machine learning methods is presented, including traditional and deep learning approaches. The proposed model yielded an F1-score value of 0.802, representing a 14% increase compared to previous methods. Finally, results from an ablation study and post-training quantization evaluation are also reported.
... However, the split ratio can influence model performance, and multiple splits have been shown to enhance reliability [31]. In contrast, a five-fold CV partitions the data into five equal subsets, rotating the training and testing roles across all folds [32]. This method minimizes overfitting and provides a more robust performance estimate [33]. ...
... Despite the strengths of these approaches, challenges in model selection remain. As Cawley and Talbot [32] emphasized, biased performance evaluation can occur when selection criteria are improperly optimized, underscoring the need for careful model selection practices. The study evaluated six ML models for behavior classification, each employing distinct learning approaches: Perceptron adjusts weights based on errors, Logistic Regression predicts probabilities using a sigmoid function, Support Vector Machine (SVM) separates classes with optimal margins [31], K-Nearest Neighbors (KNN) classifies based on the nearest data points, Random Forest (RF) aggregates multiple decision trees for robust predictions, and XGBoost (XGB) enhances accuracy through gradient-boosted trees [32]. ...
... As Cawley and Talbot [32] emphasized, biased performance evaluation can occur when selection criteria are improperly optimized, underscoring the need for careful model selection practices. The study evaluated six ML models for behavior classification, each employing distinct learning approaches: Perceptron adjusts weights based on errors, Logistic Regression predicts probabilities using a sigmoid function, Support Vector Machine (SVM) separates classes with optimal margins [31], K-Nearest Neighbors (KNN) classifies based on the nearest data points, Random Forest (RF) aggregates multiple decision trees for robust predictions, and XGBoost (XGB) enhances accuracy through gradient-boosted trees [32]. To ensure a fair and reliable assessment of both methods and their impact on classification accuracy, the study prioritized using the models with the highest accuracy in RTS as the benchmark for evaluating the CV method. ...
This study classified cows’ foraging behaviors using machine learning (ML) models evaluated through random test split (RTS) and cross-validation (CV) data partition methods. Models included Perceptron, Logistic Regression, Support Vector Machine, K-Nearest Neighbors, Random Forest (RF), and XGBoost (XGB). These models classified activity states (active vs. static), foraging behaviors (grazing (GR), resting (RE), walking (W), ruminating (RU)), posture states (standing up (SU) vs. lying down (LD)), and posture combinations with rumination and resting behaviors (RU_SU, RU_LD, RE_SU, and RE_LD). XGB achieved the highest accuracy for state classification (74.5% RTS, 74.2% CV) and foraging behavior (69.4% CV). RF outperformed XGB in other classifications, including GR, RE, and RU (62.9% CV vs. 56.4% RTS), posture (83.9% CV vs. 79.4% RTS), and behaviors-by-posture (58.8% CV vs. 56.4% RTS). Key predictors varied: speed and Actindex were crucial for GR and W when increasing and for RE and RU when decreasing. X low values were linked to RE_SU and RU_SU, while X and Z influenced RE_LD more. RTS showed higher accuracy in activity states classification while CV in foraging behaviors and by posture classification. These results emphasize CV in RF’s reliability in managing complex behavioral patterns and the importance of continuous recording devices and movement data to monitor cattle behavior accurately.
... However, the split ratio can influence model performance, and multiple splits have been shown to enhance reliability [31]. In contrast, a five-fold CV partitions the data into five equal subsets, rotating the training and testing roles across all folds [32]. This method minimizes overfitting and provides a more robust performance estimate [33]. ...
... Despite the strengths of these approaches, challenges in model selection remain. As Cawley and Talbot [32] emphasized, biased performance evaluation can occur when selection criteria are improperly optimized, underscoring the need for careful model selection practices. The study evaluated six ML models for behavior classification, each employing distinct learning approaches: Perceptron adjusts weights based on errors, Logistic Regression predicts probabilities using a sigmoid function, Support Vector Machine (SVM) separates classes with optimal margins [31], K-Nearest Neighbors (KNN) classifies based on the nearest data points, Random Forest (RF) aggregates multiple decision trees for robust predictions, and XGBoost (XGB) enhances accuracy through gradient-boosted trees [32]. ...
... As Cawley and Talbot [32] emphasized, biased performance evaluation can occur when selection criteria are improperly optimized, underscoring the need for careful model selection practices. The study evaluated six ML models for behavior classification, each employing distinct learning approaches: Perceptron adjusts weights based on errors, Logistic Regression predicts probabilities using a sigmoid function, Support Vector Machine (SVM) separates classes with optimal margins [31], K-Nearest Neighbors (KNN) classifies based on the nearest data points, Random Forest (RF) aggregates multiple decision trees for robust predictions, and XGBoost (XGB) enhances accuracy through gradient-boosted trees [32]. To ensure a fair and reliable assessment of both methods and their impact on classification accuracy, the study prioritized using the models with the highest accuracy in RTS as the benchmark for evaluating the CV method. ...
The study classified cows' foraging behaviors using machine learning (ML) models evaluated through Random Test-Split (RTS) and Cross-Validation (CV). Models in-cluded Perceptron, Logistic Regression, Support Vector Machine, K-Nearest Neighbors, Random Forest (RF), and XGBoost (XGB). These models classified activity states (Active vs. Static), foraging behaviors (Grazing (GR), Resting (RE), Walking (W), Ruminating (RU)), posture states (Standing up (SU) vs. Lying down (LD)), and activity-by-posture combinations (RU_SU, RU_LD, RE_SU, RE_LD). XGB achieved the highest accuracy for state classification (74.5% RTS, 74.2% CV) and foraging behavior (69.4% CV). RF out-performed XGB in other classifications, including GR, RE, and RU (62.9% CV vs. 56.4% RTS), posture (83.9% CV vs. 79.4% RTS), and activity-by-posture (58.8% CV vs. 56.4% RTS). Key predictors varied: Speed and Actindex were crucial for GR and W when in-creasing and for RE and RU when decreasing. X low values were linked to RE_SU and RU_SU, while X and Z influenced RE_LD more. RTS showed higher accuracy in general activity states classification while CV in foraging behaviors and by posture classification. These results emphasize CV in RF's reliability in managing complex behavioral patterns and the importance of continuous recording devices and movement metrics to monitor cattle behavior accurately.
... Accelerometers mounted on collars in the neck are widely used to detect cow activity, body posture, and head movements [20][21][22][23][24]. Sensors for localization (based, for instance, on GPS antenna) are often combined with accelerometers to track animals' positions and monitor the spatial dispersion of the herd [25][26][27]. Gyroscopes and magnetometers, also mounted on the neck, provide additional information about angular movements and head orientation [28][29][30][31][32][33][34][35]. Additionally, vocalizations can be recorded by a microphone attached to the collar [36]. ...
Wearable collar technologies have become integral to the advancement of precision livestock farming, revolutionizing how dairy cattle are monitored in terms of their behaviour, health status, and productivity. These devices leverage cutting-edge sensors, including accelerometers, RFID tags, GPS receivers, microphones, gyroscopes, and magnetometers, to provide non-invasive, real-time insights that enhance animal welfare, optimize resource use, and support decision-making processes in livestock management. This systematized review focuses on analyzing the sensors integrated into collar-based systems, detailing their functionalities and applications. However, significant challenges remain, including the high energy consumption of some sensors, the need for frequent recharging, and limited parameter coverage by individual devices. Future developments must focus on integrating multiple sensor types into unified systems to provide comprehensive data on animal behaviour, health, and environmental interactions. Additionally, advancements in energy-efficient designs, longer battery life, and cost-reduction strategies are essential to enhance the practicality and accessibility of these technologies. By addressing these challenges, wearable collar systems can play a pivotal role in promoting sustainable, efficient, and responsible livestock farming, aligning with global goals for environmental and economic sustainability. This paper underscores the transformative potential of wearable collar technologies in reshaping the livestock industry and driving the adoption of innovative farming practices worldwide.
... Though many studies exist on using motion sensors in detecting and understanding cattle behaviour, many unexploited potentials exist, such as detecting more complex behaviour and multi-dimensional movements (i.e., different movements happening simultaneously). For instance, pre-existing studies using accelerometers (Peng et al., 2019;Riaboff et al., 2020) could recognize forms of movement but are limited to only one form at a given moment or variation within the movement (Li et al., 2022). Meanwhile, in reality, certain behaviour can comprise of multiple movements occurring simultaneously, e.g., ruminating while lying or standing; aggression while ruminating and standing. ...
... McLennan et al. [6] validated an automated recording system for evaluating behavioral activity levels in sheep. Current research on gait recognition algorithms primarily focuses on analyzing behavioral patterns of sheep [7][8][9] and cattle [10][11][12]. This study has drawn on their research method, utilizing gait analysis to measure activity levels. ...
Featured Application
This research introduces a novel wearable device that uses an acceleration threshold behavior recognition method to classify horse activities into three levels: low (standing), medium (walking), and high (trotting, cantering, and galloping). The recognition algorithm is directly implemented in the hardware, which horses wear during their training sessions. This device allows for the real-time analysis of horse activity levels and the accurate calculation of the time spent in each activity state. This method provides scientific data support for horse training, facilitating the optimization of training programs.
Abstract
This study demonstrated that wearable devices can distinguish between different levels of horse activity, categorized into three types based on the horse’s gaits: low activity (standing), medium activity (walking), and high activity (trotting, cantering, and galloping). Current research in activity level classification predominantly relies on deep learning techniques, known for their effectiveness but also their demand for substantial data and computational resources. This study introduces a combined acceleration threshold behavior recognition method tailored for wearable hardware devices, enabling these devices to classify the activity levels of horses directly. The approach comprises three sequential phases: first, a combined acceleration interval counting method utilizing a non-linear segmentation strategy for preliminary classification; second, a statistical analysis of the variance among these segments, coupled with multi-level threshold processing; third, a method using variance-based proximity classification for recognition. The experimental results show that the initial stage achieved an accuracy of 87.55% using interval counting, the second stage reached 90.87% with variance analysis, and the third stage achieved 91.27% through variance-based proximity classification. When all three stages are combined, the classification accuracy improves to 92.74%. Extensive testing with the Xinjiang Wild Horse Group validated the feasibility of the proposed solution and demonstrated its practical applicability in real-world scenarios.
... The integration of advanced technologies in real-world farming operations remains a significant challenge. The need for systems that are not only accurate but also user-friendly and adaptable to different farming conditions is critical for widespread adoption [10][11][12][13][14][15][16][17][18][19]. ...
The proliferation of complex diseases in livestock, such as lumpy skin disease, demands advanced diagnostic tools that can accurately classify and predict outbreaks. This study explores the integration of complex mathematical operators within deep learning classifiers to enhance their accuracy and efficiency in diagnosing lumpy skin disease. By focusing on the decision boundary complexity, which delineates different disease states in high-dimensional spaces, and employing combinatorial optimization techniques, we develop a novel framework that significantly improves classification performance. The methodology hinges on optimizing the configuration and combination of mathematical operators, such as gradient operators and higher-order derivatives, to refine feature extraction processes. This approach allows for a more nuanced understanding of the disease features that are critical for accurate classification. Using a dataset comprised of clinical and image data from infected cattle, our enhanced classifiers demonstrate a marked improvement in predictive accuracy compared to traditional deep learning models. The case study not only underscores the potential of integrating advanced mathematical concepts into deep learning but also sets a precedent for tackling similar challenges in veterinary medicine and beyond.
... However, traditional manual monitoring of cow behaviors can be costly, inefficient, and subjective to each human observer. With advancements in technologies, previous researchers have investigated the use of sensors in order to classify different cow behaviors [1], [2], [3], [4], [5], [6], [7], [8], [9]. Although the technique of using sensors to track animal behaviors has shown promising results, it can still pose some challenges. ...
... With limited resources, it is not suitable for running deep learning or machine learning model algorithms.). Furthermore, current research on behavior recognition algorithms primarily focuses on analyzing behavioral patterns of sheep [9][10][11] and cattle [12][13][14]. Therefore, in developing a new horse behavior classification algorithm, this study referenced research methodologies from other domains. ...
Analyzing horse behavior is crucial for assessing training quality, particularly in accurately identifying actions like standing, walking, and running. Research has been conducted both domestically and internationally in this regard; however, challenges persist, including reliance on single sensors, poor real-time performance, and low identification accuracy. In view of these challenges, this study investigated real-time identification of horse behavior based on wearable devices. The system, centered around a microcontroller and utilizing a 4G network as a carrier, employs multi-axis IMU sensors as input sources to perceive horse posture. The study proposed a behavior classification method based on analysis of acceleration thresholds. The method consists primarily of two sequential stages: first, the resultant acceleration interval counting method, which employs a nonlinear segmentation approach for initial behavior classification; and second, the statistical analysis of variance parameters between segments, which when coupled with multi-level threshold processing, achieves a refined classification. Experimental results indicated that the interval counting method alone achieved an accuracy of 87.55%, while for the variance analysis method alone the accuracy was 90.87%. The proposed method, comprising the two stages, reached a classification accuracy of 91.57%, underpinning its usefulness in supporting future equine research.
