" This thesis is dedicated to my parents, Erica and all my dear friends. It wouldn't be possible to finish it without all the love I have received. " Acknowledgements The author wishes to thank: (i) Prof. Luca Benini for his high-level supervision and continuous support, in that no knowledge is possible without a master to follow; (ii) Ing.tino Ruggiero for his sincere friendship; (iii) All the...

This paper proposes a method for extracting high-level rules for expository dialogue generation. The rules are extracted from dialogues that have been authored by expert dialogue writers. We examine the rules that can be extracted by this method, focusing on whether different dialogues and authors exhibit different dialogue styles.

Real-time virtual humans are less believable than hand-animated characters, particularly in the way they perform gaze. In this paper, we provide the results of an empirical study that explores an observer’s attribution of emotional state to gaze. We have taken a set of low-level gaze behaviors culled from the nonverbal behavior literature; combined these behaviors based on a dimensional model o...

ii Abstract (in Finnish) iiiin Finnish) iii

We have witnessed a decade of exploding research interest in multimedia content analysis. The goal of content analysis has been to derive automatic methods for high-level description and annotation. In this paper we will summarize the main research topics in this area and state some assumptions that we have been using all along. We will also postulate the main future trends including usage of l...

This paper analyzes the data now available in Australia’s coordinated Electronic Theses and Dissertations gateway to show the impact of high-level institutional policy decisions on population of the individual repositories. The paper shows that just like research article repositories, voluntary ETD deposition results in repositories collecting only about 15% of the available theses, whereas man...

Prior approaches to sentence compression have taken low level syntactic constraints into account in order to maintain grammaticality. We propose and successfully evaluate a more comprehensive, generalizable feature set that takes syntactic and structural relationships into account in order to sustain variable compression rates while making compressed sentences more coherent, grammatical and rea...

The use of video and audio features for automated annotation of audio-visual data is becoming widespread. A major limitation of many of the current methods is that the stored indexing features are too low-level they relate directly to properties of the data. In this work we apply a further stage of processing that associates the feature measurements with real-world objects or events. The output...