Sourav Bhattacharya defends his PhD thesis on Continuous context inference on mobile platforms on August 25th, 2014

 

M.Sc. Sourav Bhattacharya will defend his doctoral thesis Continuous context inference on mobile platforms on Monday 25th of August.2014 at 12 o'clock in the University of Helsinki Main Building, Auditorium XIV (Unioninkatu 34, 3rd floor). His opponent is Professor Antonio Krüger (German Research Center for Artificial Intelligence and Saarland University, Germany) and custos Professor Petri Myllymäki (University of Helsinki). The defense will be held English.

Continuous context inference on mobile platforms

Mobile phones have revolutionized the way we communicate today and have become a popular computing platform. Today, off-the-shelf modern smartphones readily support a rich set of on-device sensors, such as GPS, WiFi, GSM, accelerometer, gyroscope, magnetometer and NFC. These sensors can capture various aspects of the surroundings of a user in real time, unobtrusively, and at an astounding rate. Contrary to the rapid technological development in sensing, the pace of battery capacity improvement is much slower. The limited battery power available on mobile platforms thus poses a big challenge to continuous and sustained sensing.

In this thesis, we focus on developing novel methods for continuous and sustained context inference on mobile platforms. We address challenges present in real-world deployment of two popular context recognition tasks within ubiquitous computing and mobile sensing, namely, localization and activity recognition. In the first part of the thesis, we provide a new localization algorithm for mobile devices using the existing GSM communication infrastructures, and then propose a solution for energy-efficient and robust tracking on mobile devices that are equipped with sensors such as GPS, compass, and accelerometer.

In the second part of the thesis, we propose a novel sparse-coding-based activity recognition framework that mitigates the time-consuming and costly bootstrapping process of activity recognizers employing supervised learning. The framework uses a vast amount of unlabeled data to automatically learn a sensor data representation through a set of extracted characteristic patterns and generalizes well across activity domains and sensor modalities.

Availability of the dissertation

An electronic version of the doctoral dissertation is available on the e-thesis site of the University of Helsinki at http://urn.fi/URN:ISBN:%20978-951-51-0047-4.

Printed copies are available on request from Sourav Bhattacharya: sourav.bhattacharya@cs.helsinki.fi.

13.08.2014 - 16:38 Pirjo Moen
13.08.2014 - 16:34 Pirjo Moen