Apparatus and methods for detecting emotions

A method for detecting emotional status of an individual, the method comprising:

• receiving a speech specimen generated by the individual and deriving therefrom intonation information;
• and
• generating an output indication of the individual's emotional status based on said intonation information,
• wherein said intonation information includes the number of thorns in at least a portion of the speech specimen and wherein said generating step comprises of computing the individual's excitement level in a current crucial portion defined within at least one current segment based on said number of thorns and generating an output indication of said excitement level,
• wherein said intonation information also comprises lengths of plateaus within said portion,
• and wherein said excitement level comprises a decreasing function of (a) the number of thorns in at least a portion of the speech specimen and of (b) the diversity of plateau length within said portion,
• wherein said excitement level in said crucial portion is computed as a Sum of (a) a function of a ratio between calSPT and an SPT value of said current segment and of (b) a function of a ratio between calJQ and a JQ value of the current segment;
• calSPT signifying an average of m SPT values in m respective calibration segments, m signifying a number of segments of a subiect's voice recording which are used for calibration in which a profile of truth/neutral emotional state is built for the subject,
• each SPT value comprising a function of a ratio between a number of thorns detected in an individual crucial portion of a segment and the number of samples n included in the individual crucial portion,
• calJQ signifying an average of m JQ values in m respective calibration segments;
• and
• each JQ value signifying a square root of plateau diversity in an individual crucial portion of a segment.