On Development of an Estimation Model for Instantaneous Presence in Audio-visual Content