The number of available indoor location solutions has been growing, however with insufficient precision, high implementation costs or scalability limitations. As fingerprinting-based methods rely on ubiquitous information in buildings, the need for additional infrastructure is discarded. Still, the time-consuming manual process to acquire fingerprints limits their applicability in most scenarios. This paper proposes an algorithm for the automatic construction of environmental fingerprints on multi-storey buildings, leveraging the information sources available in each scenario. It relies on unlabelled crowdsourced data from users’ smartphones. With only the floor plans as input, a demand for most applications, we apply a multimodal approach that joins inertial data, local magnetic field and Wi-Fi signals to construct highly accurate fingerprints. Precise movement estimation is achieved regardless of smartphone usage through Deep Neural Networks, and the transition between floors detected from barometric data. Users’ trajectories obtained with Pedestrian Dead Reckoning techniques are partitioned into clusters with Wi-Fi measurements. Straight sections from the same cluster are then compared with subsequence Dynamic Time Warping to search for similarities. From the identified overlapping sections, a particle filter fits each trajectory into the building’s floor plans. From all successfully mapped routes, fingerprints labelled with physical locations are finally obtained. Experimental results from an office and a university building show that this solution constructs comparable fingerprints to those acquired manually, thus providing a useful tool for fingerprinting-based solutions automatic setup.

With the fast increase in the demand for location-based services and the proliferation of smartphones, the topic of indoor localization is attracting great interest. In indoor environments, users’ performed activities carry useful semantic information. These activities can then be used by indoor localization systems to confirm users’ current relative locations in a building. In this paper, we propose a deep-learning model based on a Convolutional Long Short-Term Memory (ConvLSTM) network to classify human activities within the indoor localization scenario using smartphone inertial sensor data. Results show that the proposed human activity recognition (HAR) model accurately identifies nine types of activities: not moving, walking, running, going up in an elevator, going down in an elevator, walking upstairs, walking downstairs, or going up and down a ramp. Moreover, predicted human activities were integrated within an existing indoor positioning system and evaluated in a multi-story building across several testing routes, with an average positioning error of 2.4 m. The results show that the inclusion of human activity information can reduce the overall localization error of the system and actively contribute to the better identification of floor transitions within a building. The conducted experiments demonstrated promising results and verified the effectiveness of using human activity-related information for indoor localization.

Over recent years, there has been an increase in popularity of the acquisition of dermoscopic skin lesion images using mobile devices, more specifically using the smartphone camera. The demand for self-care and telemedicine solutions requires suitable methods to guide and evaluate the acquired images’ quality in order to improve the monitoring of skin lesions. In this work, a system for automated focus assessment of dermoscopic images was developed using a feature-based machine learning approach. The system was designed to guide the user throughout the acquisition process by means of a preview image validation approach that included artifact detection and focus validation, followed by the image quality assessment of the acquired picture. This paper also introduces two different datasets, dermoscopic skin lesions and artifacts, which were collected using different mobile devices to develop and test the system. The best model for automatic preview assessment attained an overall accuracy of 77.9% while focus assessment of the acquired picture reached a global accuracy of 86.2%. These findings were validated by implementing the proposed methodology within an android application, demonstrating promising results as well as the viability of the proposed solution in a real life scenario.

The ability to locate users and estimate traffic in mobile networks is still one of the major challenges when it comes to planning and optimizing the networks. Since indoor location is not always possible or precise, having the ability to distinguish indoor from outdoor traffic can be a valuable alternative and/or improvement. In this paper, two different machine learning algorithms are presented to classify a user’s environment, whether indoor or outdoor, using only data from a Long Term Evolution (LTE) network. To test both algorithms, two different measurement campaigns were done. Both campaigns used a smartphone to gather data from the user’s side. The first measurement campaign was done across 6 different cities, ranging from small rural areas to large urban environments, while the second was only done on a large urban city. On the second campaign, Network Traces (NT) data was also collected from the network side. The first algorithm consists on a Random Forest (RF) and the second relies on a Long Short Term Memory (LSTM), thus covering both more traditional machine learning and deep learning approaches. The results varied from 0.75 to 0.91 on the F1-Score, depending on the validation strategy, showing promising results.

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.