Textless Speech Emotion Conversion

• Textless Speech Emotion Conversion Using Discrete and Decomposed Representations
• Speech emotion conversion
• modifying the perceived emotion of a speech utterance while preserving the lexical content and speaker identity
• spoken language translation task
• decomposition of the speech signal into discrete learned representations, consisting of phonetic-content units, prosodic Features, speaker, and emotion
• modify the speech content by translating the phonetic-content units to a target emotion, and then predict the prosodic Features based on these units
• speech waveform is generated by feeding the predicted representations into a neural vocoder
• beyond spectral and parametric changes of the signal
• model non-verbal vocalizations, such as laughter insertion, yawning removal, etc