Italian CARINI  DB (narrative style)    top.gif (983 bytes)

The Italian CARINI  DB (narrative style) comprises the recordings of three novels by Dino Buzzati (“Il Colombre”, “I sette messaggeri” and “La giacca stregata”) read by a professional speaker (Claudio CARINI). The domain of this DB is a novel-like narrative reading style that is a relatively calm., clear and relaxed style. The total duration of the DB is aroung 1 hour for a total of 698 sentences and 7709 words.

EW-AV-DB (emotional Audio/Video Data Base)    top.gif (983 bytes)
EW-AV-DB is a new collected Audio Visual DB which is used to model the articulatory movements associated with emotional AV speech in Italian and to train the statistical models forn modeling the dynamics of emotional facial expressions. It’s divided in:

  • MIC-ART Emotion 1
    Italian 3 emotional isolated words (aba, ava, mamma) + 1 sentence (la mamma mangia), 6+1 emotions , 1 male actor, 5 repetitions, microphonic signal, 16 bit PCM, 16 kHz, 28 ELITE markers 16-bit PCM coded at 100 Hz (10 ms), segmented and labelled (ASCII)

  • MIC-ART Emotion 2
    Italian 7 emotional isolated nonsense words (aba, ada, aLA, adZa, ala, ana, ava) + 1 sentence (“Il fabbro lavora con forza usando il martello e la tenaglia” – lit. “the smith works with strength using the hammer and the pincer”), Such utterances provide a good phonetic coverage and they also cover the seven basic viseme classes for Italian. Each utterance was acted with seven emotional states, corresponding to the Eckman’s set – Anger, Disgust, Fear, Happiness, Sadness, Surprise - [Ekman, 1992], with the additional state ‘Neutral’. Each emotional state, except ‘Neutral’, was acted with different intensities: Low, Medium and High. Technical Details: 6+1 emotions (3 levels: low, medium high), 1 male actor, 5 repetitions, microphonic signal, 16 bit PCM, 16 kHz, 28 ELITE markers 16-bit PCM coded at 100 Hz (10 ms), segmented and labelled (ASCII: PHN, "narrow" phonetic labelling - WRD: orthographic transcription)

  • MIC-ART Emotion 3
    Provides data pertaining to the dynamics of emotional behaviour: the actor was asked to play concatenated words (short words in pairs) each with three different emotional states – anger, happiness, and neutral) at a medium intensity. In order to assess cross-language/cross-cultural effects, a set of non-sense words (abba, adda, alla, anna, avva), common to Italian and Swedish, acted with three different emotional states (Anger, Happiness, and Neutral), was also collected.

The recording procedure was the following:
short words and sentences to be played by the actor, followed by the corresponding emotional state and intensity, were announced by a speaker, according to the following scheme: <utterance/short word><emotional state><intensity> For example: Abba, happy, low. The sequences of “utterance/short word, emotional state, intensity”, covering all possible combinations for several repetitions, to be played, were generated randomly. To ensure a easier detection of the starting and ending point of the emotional behaviour, before and after the utterance/short word played emotionally, the actor delivered an additional word, respectively “chiudo” (lit. I close) and “punto” (point). For example: “CHIUDO” [abba]Hap, Low “PUNTO”

Italian ECARINI DB (emotional)    top.gif (983 bytes)

In order to collect the necessary amount of emotional speech data to train the TTS prosodic models, a professional actor (Claudio CARINI) was asked to produce vocal expressions of emotion (often using standard verbal content) as based on emotion labels and/or typical scenarios. The Emotional-CARINI (E-Carini) database recorded for this study contains the recording of a novel (“Il Colombre” by Dino Buzzati) read and acted in different elicited emotions. According to the Ekman’s theory six basic emotions, plus a neutral one, have been taken into consideration: anger, disgust, fear, happiness, sadness, and surprise. The duration of the database is about 15 minutes for each emotion.

For more information please contact :

Piero Cosi Istituto di Scienze e Tecnologie della Cognizione - Sezione di Padova "Fonetica e Dialettologia"
CNR di Padova (e-mail:


working.gif (1843 bytes)