Despite the importance of maintaining an adequate hydration status, water intake is frequently neglected due to the fast pace of people’s lives. For the elderly, poor water intake can be even more concerning, not only due to the damaging impact of dehydration, but also since seniors’ hydration regulation mechanisms tend to be less efficient. This work focuses on the recognition of the pre-drinking hand-to-mouth movement (a drink trigger) with two main objectives: predict the occurrence of drinking events in real-time and free-living conditions, and assess the potential of using this method to trigger an external component for estimating the amount of fluid intake. This shall contribute towards the efficiency of more robust multimodal approaches addressing the problem of water intake monitoring. The system, based on a single inertial measurement unit placed on the forearm, is unobtrusive, user-independent, and lightweight enough for real-time mobile processing. Drinking events outside meal periods were detected with an F-score of 97% in an offline validation with data from 12 users, and 85% in a real-time free-living validation with five other subjects, using a random forest classifier. Our results also reveal that the algorithm first detects the hand-to-mouth movement 0.70 s before the occurrence of the actual sip of the drink, proving that this approach can have further applications and enable more robust and complete fluid intake monitoring solutions.

The increasingly aging society in developed countries has raised attention to the role of technology in seniors’ lives, namely concerning isolation-related issues. Independent seniors that live alone frequently neglect meals, hydration and proper medication-taking behavior. This work aims at eating and drinking recognition in free-living conditions for triggering smart reminders to autonomously living seniors, keeping system design considerations, namely usability and senior-acceptance criteria, in the loop. To that end, we conceived a new dataset featuring accelerometer and gyroscope wrist data to conduct the experiments. We assessed the performance of a single multi-class classification model when compared against several binary classification models, one for each activity of interest (eating vs. non-eating; drinking vs. non-drinking). Binary classification models performed consistently better for all tested classifiers (k-NN, Naive Bayes, Decision Tree, Multilayer Perceptron, Random Forests, HMM). This evidence supported the proposal of a semi-hierarchical activity recognition algorithm that enabled the implementation of two distinct data stream segmentation techniques, the customization of the classification models of each activity of interest and the establishment of a set of restrictions to apply on top of the classification output, based on daily evidence. An F1-score of 97% was finally attained for the simultaneous recognition of eating and drinking in an all-day acquisition from one young user, and 93% in a test set with 31 h of data from 5 different unseen users, 2 of which were seniors. These results were deemed very promising towards solving the problem of food and fluids intake monitoring with practical systems which shall maximize user-acceptance.

The past years have witnessed a boost in fall detection-related research works, disclosing an extensive number of methodologies built upon similar principles but addressing particular use-cases. These use-cases frequently motivate algorithm fine-tuning, making the modelling stage a time and effort consuming process. This work contributes towards understanding the impact of several of the most frequent requirements for wearable-based fall detection solutions in their performance (usage positions, learning model, rate). We introduce a new machine learning pipeline, trained with a proprietary dataset, with a customisable modelling stage which enabled the assessment of performance over each combination of custom parameters. Finally, we benchmark a model deployed by our framework using the UMAFall dataset, achieving state-of-the-art results with an F1-score of 84.6% for the classification of the entire dataset, which included an unseen usage position (ankle), considering a sampling rate of 10 Hz and a Random Forest classifier.

Missing marker information is a common problem in Motion Capture (MoCap) systems. Commercial MoCap software provides several methods for reconstructing incomplete marker trajectories; however, these methods still rely on manual intervention. Current alternatives proposed in the literature still present drawbacks that prevent their widespread adoption. The lack of fully automated and universal solutions for gap filling is still a reality. We propose an automatic frame-wise gap filling routine that simultaneously explores restrictions between markers’ distance and markers’ dynamics in a least-squares minimization problem. This algorithm constitutes the main contribution of our work by simultaneously overcoming several limitations of previous methods that include not requiring manual intervention, prior training or training data; not requiring information about the skeleton or a dedicated calibration trial and by being able to reconstruct all gaps, even if these are located in the initial and final frames of a trajectory. We tested our approach in a set of artificially generated gaps, using the full body marker set, and compared the results with three methods available in commercial MoCap software: spline, pattern and rigid body fill. Our method achieved the best overall performance, presenting lower reconstruction errors in all tested conditions.

The increasing popularity of water sports—surfing, in particular—has been raising attention to its yet immature technology market. While several available solutions aim to characterise surf session events, this can still be considered an open issue, due to the low performance, unavailability, obtrusiveness and/or lack of validation of existing systems. In this work, we propose a novel method for wave, paddle, sprint paddle, dive, lay, and sit events detection in the context of a surf session, which enables its entire profiling with 88.1% accuracy for the combined detection of all events. In particular, waves, the most important surf event, were detected with second precision with an accuracy of 90.3%. When measuring the number of missed and misdetected wave events, out of the entire universe of 327 annotated waves, wave detection performance achieved 97.5% precision and 94.2% recall. These findings verify the precision, validity and thoroughness of the proposed solution in constituting a complete surf session profiling system, suitable for real-time implementation and with market potential.

Surf is an increasingly popular sport, with its evaluation being mostly qualitative and subjective. Surf analysis is often performed with commercially available systems. However, most of them are based on Global Positioning System (GPS) data and/or seem to lack in the precision and validity that surf practitioners need. Thus, a novel and accurate method to detect and characterize all surf-related events during a surf session is presented, recurring to commonly available sensors present in smartphones. Identification and characterization of stationary events, such as sitting or lying when waiting for a wave, paddle and wave riding during a whole surf session was assessed recurring to inertial data alone, GPS data alone or a combination of both. The most accurate results were obtained using a combination of both sources of data with an overall event detection accuracy of 95%, encompassing all aforementioned events. In particular, wave detection showed quite promising results with 97% for both precision and recall. These findings may indicate that the proposed solution may be suitable for the creation of a precise and valid surfer performance monitor system that could be used in real-time, providing important feedback to surf practitioners.