Learning behavioural patterns in a mobile context using smartphones
|Abstract||This thesis is focused on presenting a system for conversation and speaker detection based on the audio data collected by the smartphone devices. The system allows to detect whether two people are participants of the same conversation and state who was speaking when and for how long. The approach is privacy preserving so it is not possible to state what was said during the conversation.|
The system is based on already existing solution but it presents a more practical approach. It uses audio data recorded by the smartphone's built-in microphone as a source of information to analyze. Smartphone application is responsible for the voiced frames detection and it is used as an inteface for displaying conversations' information to the user. Server side web services collect audio features data sent by the smartphone prototype and they perform conversation and speaker detection based on comparison of a data obtained from dierent smartphones. They also provide detected conversations' data so it can be obtained by the smartphone application.
Firstly audio data analyzing tool has been created in order to facilitate calculations correctness verification of each part of the process. Afterwards smartphone application and server side web services have been implemented. All parts of the process have been thoroughly tested and the results have been analyzed.
|Type||Master's thesis [Academic thesis]|
|Publisher||Technical University of Denmark, DTU Informatics, E-mail: email@example.com|
|Address||Asmussens Alle, Building 305, DK-2800 Kgs. Lyngby, Denmark|
|Note||Supervised by Associate Professor Jakob Eg Larsen, firstname.lastname@example.org, Michael Kai Petersen and Sune Lehmann JÝrgensen, DTU Informatics|
|BibTeX data|| [bibtex]|
|IMM Group(s)||Intelligent Signal Processing|