the
∞
.ai
Search
/
Concepts
trail ·
0 nodes
Every concept
Synthesis
8 concepts, 8 reviewed and merged.
Architecture
Vocoder
A spectrogram has thrown away the phase, so turning one back into sound is not an inverse transform — it is a…
verified
Control
Prosody Control
Prosody is the rhythm, stress and intonation that make speech mean something beyond its words.
verified
Evaluation
Synthesis Evaluation
There is no word error rate for how good speech sounds.
verified
Generation
Audio Generation
Generating music or sound effects rather than speech.
verified
Methods
Voice Conversion
Keep what was said and change who said it.
verified
Representation
Speaker Embedding
A fixed-length vector that captures who is speaking rather than what they said.
verified
Speech Tokenization
Once `neural-audio-codec` turns audio into discrete tokens, speech becomes a sequence-modelling problem and…
verified
Serving
Streaming Synthesis
For a conversation, speech has to start before the sentence is finished.
verified