Jean Vanderdonckt

User interface aesthetics, a particular sub-characteristic of the ISO 25010 software quality model, is correlated to the perceived or actual usability of a graphical user interface, its user experience, and trust. While many measures, such as balance, symmetry, proportion, alignment, regularity, and simplicity, can be computed, no consensus exists today on which measure to adopt, which formula to compute for each measure, and which interpretation to give for each computed formula. To accommodate these variations and to make the assessment explicit and interpretable, we present a quality model-based approach for measuring aesthetics that defines a quality function in terms of the formula of the measures, their weights, their composition, and their overall computation. This quality model is transformed into a configuration used by GRACE, a web application developed for this purpose. When the measures, their formula, their weights, or the quality function change, the quality model changes, which is re-computed to compare it with any other model, thus making the measurement process explicit and interpretable. We apply this approach to a large dataset "Lab in the Wild'' to investigate the correlations between aesthetic measures and perceived visual complexity and colorfulness. We discuss limitations through threats to validity.

Radar sensing technologies offer several advantages over other gesture input modalities, such as the ability to reliably sense human movements, a reasonable deployment cost, insensitivity to ambient conditions such as light, temperature, and the ability to preserve anonymity. These advantages come at the price of high processing complexity mainly due to the spatio-temporal variations of gesture articulation performed by different people. Deep learning methods, such as CNN-LSTM and 3D CNN-LSTM, have a high potential to recognize radar-based gestures but usually require hundreds or thousands of labeled training samples and high processing power. Asking a lot of people to acquire a lot of gestures is particularly tedious and tiring to the point of being unrealistic. To overcome these challenges, we propose FORTE, a hand gesture recognition with few samples based on an optimized CNN architecture working on pre-processed raw data. Using a k=5-fold cross-validation, we define and compare three alternative CNNs for recognizing hand gestures acquired in a semi-mobile context of use with a portable radar attached to a smartphone. The best CNN reaches an accuracy of 94.96% with a precision of 95.92% and a recall of 96.03% for a dataset composed of solely 5 participants producing 2 samples for 20 classes covering 1 pointing, 2 pantomimic, 3 iconic, and 14 semaphoric gestures. We suggest some implications for designing radar-based gestures and we discuss the limitations of this approach.