Book excerpt: As technology continues to improve, hand gesture recognition as a form of humancomputer interaction is becoming more and more feasible. One such piece of technology, the Leap Motion Controller, provides 3D tracking data of the hands through an easy-to-use API. This thesis presents an application that uses Leap Motion tracking data to learn and recognize static and dynamic hand gestures. Gestures are recognized using statistical pattern recognition. Each gesture is defined by a set of features including fingertip positions, hand orientation, and movement. Given sufficient training data, the similarity between two gestures is measured by comparing each of their corresponding features. Two separate implementations are presented for dealing with the temporal aspect of dynamic gestures. Users are able to interact with the system using a convenient graphical user interface. The accuracy of the system was experimentally tested with the help of two separate test participants: one for the training phase and one for the recognition phase. All test gestures (both static and dynamic) were successfully recognized with minimal training data. In some cases, additional gestures were mistakenly recognized.