We illustrate a system performing multimodal human emotion detection from video input through the integration of audio emotional recognition, text emotional recognition, facial emotional recognition, and emotional recognition from a spectrogram. The outcomes of the four emotion recognition modalities are compared, and a final evaluation provides the most likely perceived emotion. The system has been devised to be easily implemented on cheap mini-computer based boards. It is conceived to be used as auxiliary tool in the field of telemedicine to remotely monitor the mood of patients and observe their healing process, which is closely related to their emotional condition.
Multimodal Mood Recognition for Assistive Scenarios
Agnese Augello;Ignazio Infantino;Giovanni Pilato;Gianpaolo Vitale
2022
Abstract
We illustrate a system performing multimodal human emotion detection from video input through the integration of audio emotional recognition, text emotional recognition, facial emotional recognition, and emotional recognition from a spectrogram. The outcomes of the four emotion recognition modalities are compared, and a final evaluation provides the most likely perceived emotion. The system has been devised to be easily implemented on cheap mini-computer based boards. It is conceived to be used as auxiliary tool in the field of telemedicine to remotely monitor the mood of patients and observe their healing process, which is closely related to their emotional condition.| File | Dimensione | Formato | |
|---|---|---|---|
|
prod_471226-doc_191333.pdf
solo utenti autorizzati
Descrizione: Multimodal Mood Recognition for Assistive Scenarios
Tipologia:
Versione Editoriale (PDF)
Licenza:
Nessuna licenza dichiarata (non attribuibile a prodotti successivi al 2023)
Dimensione
4.73 MB
Formato
Adobe PDF
|
4.73 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
