2D Stroke-gesture Recognition in Multiple Datasets

The many ways that people interact with computers include using gestures as an interface, which has become increasingly essential. There are many algorithms that are referred to as "recognizers" that are used to evaluate these gestures. The challenge, however, is in determining which recognizer is the finest because there are numerous factors to take into account rather than just one. These elements may include training time, recognition time, and recognition rate. The main objective of this is to evaluate these gesture using different dataset that is provided to determine which is the best and worst recognizer. To do this we have to take few steps in this thesis: (1) Understanding the Gesture:an evaluation tool, (2) Evaluating and comparing the results obtained from the tool, (3) Final decision table which compares and mentions which recognizers is the best and the worst.