Parkinson's Disease is one of the most common neurodegenerative disorders of the central nervous system that affects elderly. There are six main symptoms: tremors, rigidity, bradykinesia (slow movements), hand asymmetry, posture instability and freezing of gait. Nowadays any type of diagnose for this disorder is done through observation by a health care professional specialized in this area. Therefore a simpler and more efficient method that General Practioners can use to have some grounded information to decide to forward a possible patient to a specialist is needed. With this in mind different systems were studied coming to the conclusion that a mobile application is among the best options. This work can be split in four important phases (see Figure 1): (1) study of the current market for this problem and for the solution to be developed, (2) development of a smartphone application capable of gathering data of the early symptoms of Parkinson's taking into consideration all the smartphone's specifications; (3) use the application to gather data from real patients and a control group and (4) test and select a classification algorithm. The first phase involved two research topics: problem and solution. The problem consisted in studying all the symptoms that could theoretically be detected by the different smartphone components. The solution consisted in studying the different methods used to solve such a problem using data mining techniques (different feature selection and classification algorithms that best take advantage of the nature of the data gathered). The second phase consisted in the development of the smartphone application with four components (spiral analysis, tap analysis, simple questions and gait analysis). The third phase was dedicated in building the control group gathering data from healthy people and a Parkinson patients group for a total of 35 subjects. Finally, the fourth phase was using the studied algorithms to filter the different features- and compare the different algorithms selected. With the available data from the test subjects it was possible to achieve promising results from the gait analysis of the patients where the pelvic sway was a good feature to help differentiate Parkinson patients from healthy ones

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.

No abstract

The ongoing reading process of digital meters is time-consuming and prone to errors, as operators capture images and manually update the system with the new readings. This work proposes to automate this operation through a deep learning-powered solution for universal controllers and flow meters that can be seamlessly incorporated into operators’ existing workflow. Firstly, the digital display area of the equipment is extracted with a screen detection module, and a perspective correction step is performed. Subsequently, the text regions are identified with a fine-tuned EAST text detector, and the important readings are selected through template matching based on the expected graphical structure. Finally, a fine-tuned convolutional recurrent neural network model recognizes the text and registers it. Evaluation experiments confirm the robustness and potential for workload reduction of the proposed system, which correctly extracts 55.47% and 63.70% of the values for reading in universal controllers, and 73.08% of the values from flow meters. Furthermore, this pipeline performs in real time in a low-end mobile device, with an average execution time in preview of under 250 ms for screen detection and on an acquired photo of 1500 ms for the entire pipeline.

No abstract