—The aim of the paper was to analyze how to acquire and process EEG data with a simplified, commercially applicable EEG interface and to check whether it is possible to recognize human emotions with it. The EEG data gathering station was built and the data was gathered from the subjects. Then, the data was processed to apply it to the classifier training. The AutoML software was used to find the best ML model, and it was also built manually to prove the output accuracy was reliable and there was no overfitting. The AutoML experiment has shown that the best classifier was the boosted decision tree algorithm, and building it manually resulted in an accuracy of recognizing four distinct emotions equal to 99.80%.
The aim of this work was to study whether word prediction can be applied to virtual Braille keyboards and can improve typing text by visually impaired smartphone user. First the keyboards’ advantages and disadvantages were compared to choose one to extend with word prediction mechanism. Next the method was proposed and implemented in a form of a mobile application. Due to Braille code’s structure, it was possible to apply not only word, but also a letter and dot prediction. Finally, both number of dots and letters necessary for predicting letters and words respectively within the shortest time were studied. This work verified that prediction in on-screen Braille keyboards is possible and brings noticeable benefits in typing speed. The most important observation is that just after the first dot (or blank place) it is worth searching the suggested letters and words. Keywords–Text entry; Braille code; Letter prediction; Word prediction; Virtual keyboard.
—Various kinds of vibrotactile information have been recorded from real textures and used to present high-quality tactile sensations via tactile displays. However, it is unrealistic to collect large amounts of vibrotactile data under many different conditions. Thus, we develop a method whereby recorded data can be changed to represent conditions differing from those at the time of initial recording. In the first step, we construct a data generation model using a Generative Adversarial Network (GAN). The model makes simple calculations and generates un- known data from recorded acceleration data obtained by rubbing real objects. The model can generate three-axis, time-series data. To evaluate the quality of the data generated, we devised a string-based tactile display and presented generated vibrotactile information to users. Users reported that the generated data were indistinguishable from real data.
This work is supported by the Spanish MiNeCo underprojects HuMoUR TIN2017-90086-R and Maria de MaeztuSeal of Excellence MDM-2016-0656. We also thank MartaAltarriba Fatsini for her support derivating the formulas tocompute skeleton angle rotations.
For the deaf and hard-of-hearing to be able to go out safely, they must be able to recognize alarm sounds (horns, bicycle bells, ambulance sirens, etc.) among various environmental sounds. Therefore, it is crucial to be able to transmit these kinds of sounds to such people, even in noisy environmental conditions. In this paper, we propose and develop an alarm sound classification system using deep neural networks. The system works on smartphones that can always be carried by the users when they are going out. Besides, we performed evaluation experiments to verify the effectiveness of the system using the 5-fold cross-validation method. Furthermore, we evaluate the classification rate for unlearned data and re-evaluate one by adding data downloaded from the web. We also discuss the limitations of the system to improve it and make it more useful. Keywords–Alarm sound; Classification; Deaf and hard-ofhearing; Neural network; Smartphone.
— Immersive Virtual Reality (IVR) may potentially effect considerable lifestyle changes in societies, comparable to those seen with the spread of smartphones. Questions arise as to the significance of IVR, and how people will respond to this type of innovation. The article presents the results of a qualitative study which assesses the reactions of adults from Generations X, Y and Z to IVR. 18 people aged 20-55 took part in the study; seven IVR applications were used. The study assessed participants' reactions, level of presence, affective response and susceptibility to cybersickness. The development potential of IVR was also considered. It was assumed that older generations would be less present in the IVR and their subjective assessment of satisfaction would be lower. The results of the study confirmed the hypothesis that, as people age, their level of presence in IVR decreases, but surprisingly, it emerged that satisfaction with being in IVR increases along with the age of the participants.
