Abstract
Over 250 million people, worldwide, are affected by chronic lung conditions such as Asthma and COPD. These can cause breathlessness, a harsh decrease in quality of life and, if left undetected or not properly managed, even death. In this paper, we approached part of the lines of development suggested upon earlier work. This concerned the development of a system design for a smartphone lung function classification app, which would only use recordings from the built-in microphone. A more systematic method to evaluate the relevant combinations of methods was devised and an additional set of 44 recordings was used for testing purposes. The previous 101 were kept for training the models. The results enabled to further reduce the signal processing pipeline leading to the use of 6 envelopes, per recording, half of the previous amount. An analysis of the classification performances is provided for both previous tasks: differentiation into Normal from Abnormal lung function, and between multiple lung function patterns. The results from this project encourage further development of the system.

Abstract
Worldwide, over 250 million people are affected by chronic lung conditions such as Asthma and COPD. These can cause breathlessness, a harsh decrease in quality of life and, if not detected and duly managed, even death. In this paper, we aim to find the best and most efficient combination of signal processing and machine learning approaches to produce a smartphone application that could accurately classify lung function, using microphone recordings as the only input. A total of 61 patients performed the forced expiration maneuver providing a dataset of 101 recordings. The signal processing comparison experiments were conducted in a backward selection approach, reducing from 54 to 12 final envelopes, per recording. The classification experiments focused first on differentiating Normal from Abnormal lung function, and second in multiple lung function patterns. The results from this project encourage further development of the system.

Abstract
Nowadays, there are several communication gaps that isolate deaf people in several social activities. This work studies the expressiveness of gestures in Portuguese Sign Language (PSL) speakers and their differences between deaf and hearing people. It is a first effort towards the ultimate goal of understanding emotional and behaviour patterns among such populations. In particular, our work designs solutions for the following problems: (i) differentiation between deaf and hearing people, (ii) identification of different conversational topics based on body expressiveness, (iii) identification of different levels of mastery of PSL speakers through feature analysis. With these aims, we build up a complete and novel dataset that reveals the duo-interaction between deaf and hearing people under several conversational topics. Results show high recognition and classification rates.

Abstract
This paper presents the results of an experimental study of the popular Bag-of-Features (BoF) model for the application of unsupervised learning of images, or image clustering. Although this method has been extensively applied for image classification and scene recognition, there has been few works which employ it in an unsupervised way. Also, due to the fact that the BoF model requires a great amount of steps, algorithms and parameter settings, we felt like there was a lack of detailed studies about the subject. We implemented testing routines in Python which we made publicly available in GitHub. In order to assess the performance of the model, three image datasets were used, namely, Coil-20 dataset, Natural and Urban dataset and Event dataset. The results obtained indicate that the BoF method provides a good representation of simple image collections for the purpose of clustering. However, it requires fine tunning of the parameters and algorithms for each dataset and obtains poor results for more complex scene datasets. We can therefore conclude that more advanced techniques are required in order to be able to effectively extract information from large image collections.

Abstract
ENDURE and BLUECOM+ are two EEA Grants projects that aim at enabling the monitoring of large and remote ocean areas by providing wireless communications and energy to humans and systems, namely unmanned vehicles, which are crucial for making large scale ocean monitoring cost-effective. ENDURE targets enabling autonomous underwater vehicles (AUVs) to remain in operation for longer periods of time than what is practical today, thus increasing the possibility of covering larger areas at lower costs. BLUECOM+ aims at connecting systems and humans in remote ocean areas by providing cost-effective, broadband, and reliable communications in alternative to satellite communications. Together these projects will bring up the enablers for data collection under the environmental monitoring programme, as well as to the achievement and sustainability of a GES in marine waters. © 2016 Taylor & Francis Group, London.

Abstract
Excessive dead time in complementary switches causes significant energy losses in DC-DC power conversion. The optimization of dead time prevents the degradation of overall efficiency by minimizing the body diode conduction of power switches and, as a consequence, also reduces reverse recovery losses. The present work aims at analyzing the influence of one of the most important characteristics of a digital controller, the timer resolution, in the context of dead-time optimization for synchronous buck converters. In specific, the analysis quantifies the efficiency dependency on the timer resolution, in a parameter set that comprises duty-cycle and dead-time, and also converter frequency and analog-to-digital converter accuracy. Based on a sensorless optimization strategy, the relationship between all these limiting factors is described, such as the number of bits of timer and analog-to-digital converter. To validate our approach experimental results are provided using a 12-to-1.8V DC-DC converter, controlled by low- and high-resolution pulse-width modulation signals generated with an XMC4200 microcontroller from Infineon Technologies. The measured results are consistent with our analysis, which predicts the power efficiency improvements not only with a fixed dead time approach, but also with the increment of timer resolution.