@article{wu_nichols_foster_martin_dieffenderfer_enomoto_lascelles_russenberger_brenninkmeyer_bozkurt_et al._2023, title={An Exploration of Machine Learning Methods for Gait Analysis of Potential Guide Dogs}, url={https://doi.org/10.1145/3637882.3637883}, DOI={10.1145/3637882.3637883}, abstractNote={Gait analysis is important for guide dog organizations, as ideal guide dogs have a smooth and efficient gait, where they can also easily shift between and maintain various gaits. Gait quality and natural traveling speed are two of the multiple factors important in matching a guide dog to its visually impaired handler. Gait evaluation typically includes subjective visual observation of the dog or objective assessments obtained from special-designed equipment. Guide dog organizations need a method to easily collect and analyze objective data of gait information. In this work, we explored how various machine learning models could learn and analyze gait patterns from inertial measurements data that were collected during two different data collection experiments using a wearable sensor device. We also evaluated how well each machine learning model could generalize behavior patterns from various dogs under different environments. Additionally, we compared how sensor placement locations could affect gait prediction performance by attaching the sensor device to the dog’s neck and back area respectively. The tested machine learning models were able to classify different gaits in the range of 42% to 91% in terms of accuracy, and predict various gait parameters with an error rate ranging from 14% to 29% depending on the setup. Furthermore, we also observed that using behavior data collected from the neck region contains more movement information than the back area. By performing a cross-dataset generalization test on the machine learning models, we found that even with performance drop, the models were able to learn gait-specific behavior patterns that are generalizable for different dogs. Although the results were preliminary, the proposed gait analysis exploration still showed promising potential for studying behavior patterns of candidate guide dogs.}, journal={TENTH INTERNATIONAL CONFERENCE ON ANIMAL-COMPUTER INTERACTION, ACI 2023}, author={Wu, Yifan and Nichols, Colt and Foster, Marc and Martin, Devon and Dieffenderfer, James and Enomoto, Masataka and Lascelles, B. Duncan X. and Russenberger, Jane and Brenninkmeyer, Gerald and Bozkurt, Alper and et al.}, year={2023} }
 @article{wang_foster_bozkurt_roberts_2022, title={Motion-Resilient ECG Signal Reconstruction from a Wearable IMU through Attention Mechanism and Contrastive Learning}, url={https://doi.org/10.1145/3565995.3566037}, DOI={10.1145/3565995.3566037}, abstractNote={Wearable electrocardiogram (ECG) sensors can detect dogs’ heartbeat signals and have proven useful in monitoring dogs’ welfare and predicting temperament scores in structured evaluations of potential guide dog puppies. Despite advances in the ergonomics, performance, and usability of ECG sensor technologies specifically designed for dogs, deploying those systems in the real world imposes challenges such as training human operators to ensure electrodes’ proper contact with the skin and, especially in the case of puppies, socialization to achieve comfort and reduce behavioral inhibition. Seismocardiogram signal is an alternate modality for heartbeat signals and is acquired using the Inertial Measurement Unit (IMU), which is commercially available, widely deployed, and does not require skin-contact. However, the extracted signals from IMU are subject to heavy influences from motion and other noise sources. In this paper, we present a method that enables extracting the similar physiological parameters ECG provides using easier-to-deploy IMU sensors. We propose and evaluate a machine learning framework that reconstructs ECG signals from IMU signals even under moderate to heavy movements. Our study investigated two artificial neural network architectures to overcome severe noise artifacts in the IMU signal resulting from dogs’ movements and environmental factors. The first architecture combines the attention mechanism and convolution layers to extract important features from the temporal IMU input. The second architecture adapts contrastive representation learning to the regression problem and learns a more effective embedding for the ECG reconstruction. The qualitative inspection and quantitative analysis based on F1 scores of the R-peak alignment demonstrate the effectiveness of the two proposed models in removing motion noises and reconstructing realistic ECG signals, achieving an F1 score of 0.72 in the best case compared to 0.29 from the baseline.}, journal={NINTH INTERNATIONAL CONFERENCE ON ANIMAL-COMPUTER INTERACTION, ACI 2022}, author={Wang, Jianxun and Foster, Marc and Bozkurt, Alper and Roberts, David L.}, year={2022} }
 @article{foster_wu_roberts_bozkurt_2022, title={Preliminary Evaluation of a System with On-Body and Aerial Sensors for Monitoring Working Dogs}, volume={22}, ISSN={["1424-8220"]}, url={https://doi.org/10.3390/s22197631}, DOI={10.3390/s22197631}, abstractNote={This paper presents a system for behavioral, environmental, and physiological monitoring of working dogs using on-body and aerial sensors. The proof of concept study presented here includes two trained dogs performing nine scent detection tasks in an uncontrolled environment encompassing approximately two acres. The dogs were outfitted with a custom designed wearable harness to monitor their heart rate, activity levels and skin temperature. We utilized a commercially available micro-air vehicle to perform aerial sensing by tracking the terrain and movement of the dog in the outdoor space. The dogs were free to explore the space working at maximal speeds to complete a scent-based search-and-retrieval task. Throughout the experiment, the harness data was transferred to a base station via Wi-Fi in real-time. In this work, we also focused on testing the performance of a custom 3D electrode with application specific ergonomic improvements and adaptive filter processing techniques to recover as much electrocardiography data as possible during high intensity motion activity. We were able to recover and use 84% of the collected data where we observed a trend of heart rate generally increasing immediately after successful target localization. For tracking the dogs in the aerial video footage, we applied a state-of-the-art deep learning algorithm designed for online object tracking. Both qualitative and quantitative tracking results are very promising. This study presents an initial effort towards deployment of on-body and aerial sensors to monitor the working dogs and their environments during scent detection and search and rescue tasks in order to ensure their welfare, enable novel dog-machine interfaces, and allow for higher success rate of remote and automated task performance.}, number={19}, journal={SENSORS}, author={Foster, Marc and Wu, Tianfu and Roberts, David L. and Bozkurt, Alper}, year={2022}, month={Oct} }
 @article{wu_holder_foster_williams_enomoto_lascelles_bozkurt_roberts_2022, title={Spatial and Temporal Analytic Pipeline for Evaluation of Potential Guide Dogs Using Location and Behavior Data}, url={https://doi.org/10.1145/3565995.3566033}, DOI={10.1145/3565995.3566033}, abstractNote={Training guide dogs for visually-impaired people is a resource-consuming task for guide dog schools. This task is further complicated by a dearth of capabilities to objectively measure and analyze candidate guide dogs’ temperaments as they are placed with volunteer raisers away from guide dog schools for months during the raising process. In this work, we demonstrate a preliminary data analysis workflow that is able to provide detailed information about candidate guide dogs’ day to day physical exercise levels and gait activities using objective environmental and behavioral data collected from a wearable collar-based Internet of Things device. We trained and tested machine learning models to analyze different gait types including walking, pacing, trotting and mixture of walk and trot. By analyzing data both spatially and temporally, a location and behavior summary for candidate dogs is generated to provide insight for guide dog training experts, so that they can more accurately and comprehensively evaluate the future success of the candidate. The preliminary analysis revealed movement patterns for different location types which reflected the behaviors of candidate guide dogs.}, journal={NINTH INTERNATIONAL CONFERENCE ON ANIMAL-COMPUTER INTERACTION, ACI 2022}, author={Wu, Yifan and Holder, Timothy and Foster, Marc and Williams, Evan and Enomoto, Masataka and Lascelles, B. Duncan X. and Bozkurt, Alper and Roberts, David L.}, year={2022} }
 @article{ahmmed_holder_foster_castro_patel_torfs_bozkurt_2021, title={Noncontact Electrophysiology Monitoring Systems for Assessment of Canine-Human Interactions}, ISSN={["1930-0395"]}, url={http://dx.doi.org/10.1109/sensors47087.2021.9639748}, DOI={10.1109/SENSORS47087.2021.9639748}, abstractNote={Canine-assisted interactions have enormous potential in coping with psychological disorders and stress. It has been actively used for improving the mood of hospitalized patients, especially those suffering from chronic diseases like cancer. However, little progress has been made to enable the assessment of these interactions between the patient and the animal in a quantitative and undisruptive way. In this paper, we present a capacitively coupled biopotential recording system custom-designed for animal-human dyads. This system uses noncontact electrodes to monitor the heart rate and its variability to evaluate the physiological basis of the animal-assisted therapies. Preliminary in vivo evaluation of the system in humans and canines demonstrates promising measurement accuracy. The mean absolute error of the estimated heart rate was less than 0.25 BPM in reference to a commercial electrocardiography device. The future integration of this system into ergonomic form factors could enable a better understanding of animal-human interactions during canine-assisted therapy sessions by realizing an unobtrusive and continuous monitoring platform.}, journal={2021 IEEE SENSORS}, publisher={IEEE}, author={Ahmmed, Parvez and Holder, Timothy and Foster, Marc and Castro, Ivan D. and Patel, Aakash and Torfs, Tom and Bozkurt, Alper}, year={2021} }
 @article{foster_brugarolas_walker_mealin_cleghern_yuschak_clark_adin_russenberger_gruen_et al._2020, title={Preliminary Evaluation of a Wearable Sensor System for Heart Rate Assessment in Guide Dog Puppies}, volume={20}, ISSN={["1558-1748"]}, url={https://doi.org/10.1109/JSEN.2020.2986159}, DOI={10.1109/JSEN.2020.2986159}, abstractNote={This paper details the development of a novel wireless heart rate sensing system for puppies in training as guide dogs. The system includes a harness with on-board electrocardiography (ECG) front-end circuit, inertial measurement unit and a micro-computer with wireless capability where the major research focus of this paper was on the ergonomic design and evaluation of the system on puppies. The first phase of our evaluation was performed on a Labrador Retriever between 12 to 26 weeks in age as a pilot study. The longitudinal weekly data collected revealed the expected trend of a decreasing average heart rate and increased heart rate variability as the age increased. In the second phase, we improved the system ergonomics for a larger scale deployment in a guide dog school (Guiding Eyes for the Blind (Guiding Eyes)) on seventy 7.5-week-old puppies (heart rate coverage average of 86.7%). The acquired ECG based heart rate data was used to predict the performance of puppies in Guiding Eyes’s temperament test. We used the data as an input to a machine learning model which predicted two Behavior Checklist (BCL) scores as determined by expert Guiding Eyes puppy evaluators with an accuracy above 90%.}, number={16}, journal={IEEE SENSORS JOURNAL}, publisher={Institute of Electrical and Electronics Engineers (IEEE)}, author={Foster, Marc and Brugarolas, Rita and Walker, Katherine and Mealin, Sean and Cleghern, Zach and Yuschak, Sherrie and Clark, Julia Condit and Adin, Darcy and Russenberger, Jane and Gruen, Margaret and et al.}, year={2020}, pages={9449–9459} }