We are constantly surrounded by electronic interfaces: elevators, automatic doors, ticket machines, remote controls, voice recognition systems, etc. With a high degree of networking between our devices, interfaces must seamlessly integrate into ecosystems of information acquisition and processing. Thus, ergonomic sensor placement, which integrates discreetly into our accustomed environment, is gaining ground.
The course explores the possibilities of this ergonomic, embedded interaction using the example of acoustic interfaces. We will develop electronic sensor systems for the detection of vibrations (structure-borne sound) in integrated surfaces and experiment with machine learning methods to differentiate between vibration gestures (e.g. knocking noises, friction noises, etc.). We will use the Teensy 4.0 hardware platform to classify sounds in the environment. Therefore, a budget of (30-40€) should be planned for the Teensy and other components.
Along the course an introduction to C++ and Object Oriented Programming will be given, which we will directly apply to program a simple Machine Learning Library on the Teensy Hardware.
The Course will be taught in video sessions and regular assignments every two weeks. A working video and audio connection is mandatory throughout the course.
If you are interested, please send a letter of motivation and portfolio to clemens.wegener ( at ) uni-weimar (dot) de. Basic knowledge of electronic sensors will be taught in the course. Basic knowledge of programming is a requirement!