This panel shows digital pen features that encode the input signal. They can be used for analysing and classifying handwritten texts, sketches, or gestures. You can find more information in our paper pEncode: A Tool for Visualizing Pen Signal Encodings in Real-Time

This panel shows the results of the classification: the probability of the sketch belonging to a certain class. For classification, we use logistic regression: an effective algorithm which is commonly used for many binary and multiobject classification tasks. As it requires at least two classes in the training set, in case where only one class if defined we fall back to the distance based approach. That is, we output the maximal cosine similarity between the sketch (its feature vector) and known samples.