Search and Rescue (SaR) dogs are important assets in the hands of first responders, as they have the ability to locate the victim even in cases where the vision and or the sound is limited, due to their inherent talents in olfactory and auditory senses. In this work, we propose a deep-learning-assisted implementation incorporating a wearable device, a base station, a mobile application, and a cloud-based infrastructure that can first monitor in real-time the activity, the audio signals, and the location of a SaR dog, and second, recognize and alert the rescuing team whenever the SaR dog spots a victim. For this purpose, we employed deep Convolutional Neural Networks (CNN) both for the activity recognition and the sound classification, which are trained using data from inertial sensors, such as 3-axial accelerometer and gyroscope and from the wearable’s microphone, respectively. The developed deep learning models were deployed on the wearable device, while the overall proposed implementation was validated in two discrete search and rescue scenarios, managing to successfully spot the victim (i.e., obtained F1-score more than 99%) and inform the rescue team in real-time for both scenarios.

During the last decade, new technological tools have emerged to provide first responders with augmented senses in emergency and disaster situations. Some of these tools focus on providing extra information about their surroundings. However, despite augmenting first responders’ capabilities, the quantity and the way that this information is presented can affect their cognitive load. This manuscript presents an integrated framework that allows real-time biosignals monitoring to analyze physical constants and correlate them with subjective cognitive load tests. Biosignals monitoring allows alarms to be raised related to the physical status while cognitive load values will allow modulation of the amount of information that can be assimilated by the first responder in operation. In-lab and practice experimental tests have been conducted to create a fully functional framework. During the technical validation, a strong dispersion of subjective cognitive load by means of NASA-TLX questionnaires has been found between participants. Nonetheless, the developed framework allows extraction of relationships between biosignals and cognitive load, with special attention to the respiration rate and eye movements.

The widespread use of social networks has brought to the foreground a very important issue, the veracity of the information circulating within them. Many natural language processing methods have been proposed in the past to assess a post’s content with respect to its reliability; however, end-to-end approaches are not comparable in ability to human beings. To overcome this, in this paper, we propose the use of a more modular approach that produces indicators about a post’s subjectivity and the stance provided by the replies it has received to date, letting the user decide whether (s)he trusts or does not trust the provided information. To this end, we fine-tuned state-of-the-art transformer-based language models and compared their performance with previous related work on stance detection and subjectivity analysis. Finally, we discuss the obtained results.

Cultural heritage sites, apart from being the tangible link to a country’s history and culture, actively contribute to the national economy, offering a foundation upon which cultural tourism can develop. This importance at the cultural and economic level, advocates for the need for preservation of cultural heritage sites for the future generations. To this end, advanced monitoring systems harnessing the power of sensors are deployed near the sites to collect data which can fuel systems and processes aimed at protection and preservation. In this paper we present the use of acoustic sensors for safeguarding cultural sites located in rural or urban areas, based on a novel data flow framework. We developed and deployed Wireless Acoustic Sensors Networks that record audio signals, which are transferred to a modular cloud platform to be processed using an efficient deep learning algorithm (f1-score: 0.838) to identify audio sources of interest for each site, taking into account the materials the assets are made of. The extracted information is presented exploiting the designed STORM Audio Signal ontology and then fused with spatiotemporal information using semantic rules. The results of this work give valuable insight to the cultural experts and are publicly available using the Linked Open Data format.