Landmark mining on a smartwatch using speech recognition

The smartwatch is a representative of the wearables, mini-computers worn on the body, which often only require casual interaction. In this project the advantages of smartwatches should be used to add landmarks to a database as performantly as possible and provide them for further usage. The mobile terminal should therefore be operated with voice input in a clearly structured speech dialogue to make the interaction as simple and intuitive as possible. Additionally a smartphone paired with the smartwatch distributes the user's GPS-data which help to gather data for landmarks in URWalking, a research project for pedestrian navigation at the University of Regensburg. The special problem is on the one hand to find solutions for hardware-related limitations such as a very small display size. On the other hand a way for the translation of plain-language input to technically evaluable data has to be found. In a final evaluation the system should be tested under real-life circumstances.