Time-based mood recognition in film music
Music in film can be used to illustrate the action on screen; it accompanies a considerable proportion of the duration of films and reinforces a mood, creating another dimension in cinema. This establishes the importance and influence music has over communicating an emotive state. By researching the relation to mood changes in current cinematic music and musical characteristics, a better understanding of composition and mood analysis in music can be achieved. At present there is a lack of research on categorising music in film based on mood and emotion from the film narrative. However, there is general research on classifying music by mood. Current research focuses on generalised classification for music because of the increase of digital music accessible to the public and a demand for automatic organising of large music collections.
The aim of this research is to identify dynamic fluctuations of mood in current film music. It focuses on novel techniques for measuring emotive content of an audio piece over time. This research is relevant to not only film music composition but also for television, documentary and game composition. It can also be relevant for creative and artistic purposes and music collection data for musicians, composers, producers and editors.
C4DM theme affiliation:
Machine Listening, Music Cognitio