A framework for energy-efficient equine activity recognition with leg accelerometers

Eerdekens, Anniek; Deruyck, Margot; Fontaine, Jaron; Martens, Luc; Poorter, Eli De; David, Paul A.; Joseph, Wout

doi:10.1016/j.compag.2021.106020

Cited by 20 publications

(27 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The jumping horses landed almost always on the correct leg during jumping training, so only two flying changes were performed by the jumping horses. Flying changes are considered as an important jumping training activity and therefore, we added data of three dressage horses (two in the training set and one in the validation set) performing nineteen flying changes to test the model's performance.Because the data is measured at 50 Hz and the optimal sampling rate in our previous work [14] was set at 10 Hz, the dataset is sub-sampled at this rate. Figure 5 illustrates the jumping training classification results using 2 s samples of accelerometer data sampled at 10 Hz.…”

Section: Results For Jumpingmentioning

confidence: 99%

“…To this end, the max-pooling layer output of the CNN is flattened and fused with additional features. For each time window, a set of feature characteristics was extracted from the acceleration signals that already demonstrated to be important for classifying animal activities [14,34,35]. The input vector of the hybrid CNN contains raw accelerometer data samples, i.e., the acceleration of the left (L) and right (R) leg in three directions (a xL , a yL , a zL , a xR , a yR , a zR ).…”

Section: Activity Classificationmentioning

confidence: 99%

See 1 more Smart Citation

Horse Jumping and Dressage Training Activity Detection Using Accelerometer Data

Eerdekens

Deruyck

Fontaine

et al. 2021

Animals

Self Cite

View full text Add to dashboard Cite

Equine training activity detection will help to track and enhance the performance and fitness level of riders and their horses. Currently, the equestrian world is eager for a simple solution that goes beyond detecting basic gaits, yet current technologies fall short on the level of user friendliness and detection of main horse training activities. To this end, we collected leg accelerometer data of 14 well-trained horses during jumping and dressage trainings. For the first time, 6 jumping training and 25 advanced horse dressage activities are classified using specifically developed models based on a neural network. A jumping training could be classified with a high accuracy of 100 %, while a dressage training could be classified with an accuracy of 96.29%. Assigning the dressage movements to 11, 6 or 4 superclasses results in higher accuracies of 98.87%, 99.10% and 100%, respectively. Furthermore, during dressage training, the side of movement could be identified with an accuracy of 97.08%. In addition, a velocity estimation model was developed based on the measured velocities of seven horses performing the collected, working, and extended gaits during a dressage training. For the walk, trot, and canter paces, the velocities could be estimated accurately with a low root mean square error of 0.07 m/s, 0.14 m/s, and 0.42 m/s, respectively.

show abstract

Section: Results For Jumpingmentioning

confidence: 99%

Section: Activity Classificationmentioning

confidence: 99%

Horse Jumping and Dressage Training Activity Detection Using Accelerometer Data

Eerdekens

Deruyck

Fontaine

et al. 2021

Animals

Self Cite

View full text Add to dashboard Cite

show abstract

“…For example, feed-forward neural networks (FNNs) and long shortterm memory (LSTM) models were applied to automatically recognize cattle behaviors (e.g., feeding, lying, and ruminating) using data collected from inertial measurement units (IMUs) [12,13]. Convolutional neural networks (CNNs), which accurately capture local temporal dependency and scale invariance in signals, were developed in automated equine activity classification based on triaxial accelerometer and gyroscope data [1,14,15]. FilterNet, presented based on CNN and LSTM architectures, was adopted to classify important health-related canine behaviors (e.g., drinking, eating, and scratching) using a collar-mounted accelerometer [16].…”

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…The behavior of horses provides rich insight into their mental and physical status and is one of the most important indicators of their health, welfare, and subjective state [ 1 ]. However, behavioral monitoring for animals, to date, largely relies on manual observations, which are labor-intensive, time-consuming, and prone to subjective judgments of individuals [ 1 ]. The use of sensors and machine learning is well-established in monitoring gait change [ 2 ], and for lameness detection as part of the equine veterinary examination, increasing the accuracy of identification of subtle lameness, which is one of the most expensive health issues in the equine industry [ 3 , 4 ].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Cross-Modality Interaction Network for Equine Activity Recognition Using Imbalanced Multi-Modal Data

Mao

Huang

Gan

et al. 2021

Sensors

View full text Add to dashboard Cite

With the recent advances in deep learning, wearable sensors have increasingly been used in automated animal activity recognition. However, there are two major challenges in improving recognition performance—multi-modal feature fusion and imbalanced data modeling. In this study, to improve classification performance for equine activities while tackling these two challenges, we developed a cross-modality interaction network (CMI-Net) involving a dual convolution neural network architecture and a cross-modality interaction module (CMIM). The CMIM adaptively recalibrated the temporal- and axis-wise features in each modality by leveraging multi-modal information to achieve deep intermodality interaction. A class-balanced (CB) focal loss was adopted to supervise the training of CMI-Net to alleviate the class imbalance problem. Motion data was acquired from six neck-attached inertial measurement units from six horses. The CMI-Net was trained and verified with leave-one-out cross-validation. The results demonstrated that our CMI-Net outperformed the existing algorithms with high precision (79.74%), recall (79.57%), F1-score (79.02%), and accuracy (93.37%). The adoption of CB focal loss improved the performance of CMI-Net, with increases of 2.76%, 4.16%, and 3.92% in precision, recall, and F1-score, respectively. In conclusion, CMI-Net and CB focal loss effectively enhanced the equine activity classification performance using imbalanced multi-modal sensor data.

show abstract

AI‐Enabled Micro Motion Sensors for Revealing the Random Daily Activities of Caged Mice

Liu

Chen

Chan³

et al. 2023

Advanced Intelligent Systems

View full text Add to dashboard Cite

More than 120 million mice and rats are used yearly for scientific purposes. While tracking their motion behaviors has been an essential issue for the past decade, present techniques, such as video‐tracking and IMU‐tracking have considerable problems, including requiring a complex setup or relatively large IMU modules that cause stress to the animals. Here, we introduce a wireless IoT motion sensor (i.e., weighing only 2 g) that can be attached and carried by mice to collect motion data continuously for several days. We also introduce a combined segmentation method and an imbalanced learning process that are critical for enabling the recognition of common but random mouse behaviors (i.e., resting, walking, rearing, digging, eating, grooming, drinking water, and scratching) in cages with a macro‐recall of 94.55%. An interactive preprint version of the article can be found at: https://doi.org/10.22541/au.166005321.10787501/v1.

show abstract

A framework for energy-efficient equine activity recognition with leg accelerometers

Abstract: A framework for energy-efficient equine activity recognition with leg accelerometers.

Cited by 20 publications

References 35 publications

Horse Jumping and Dressage Training Activity Detection Using Accelerometer Data

Horse Jumping and Dressage Training Activity Detection Using Accelerometer Data

Cross-Modality Interaction Network for Equine Activity Recognition Using Imbalanced Multi-Modal Data

AI‐Enabled Micro Motion Sensors for Revealing the Random Daily Activities of Caged Mice

Contact Info

Product

Resources

About