Spectrographic speech mask estimation using the time-frequency correlation of speech presence.
Ge ZhanZhaoqiong HuangDongwen YingJielin PanYonghong YanPublished in: INTERSPEECH (2015)
Keyphrases
- speech recognition
- speech signal
- audio visual
- speech synthesis
- recognition engine
- vocal tract
- automatic speech recognition
- spoken language
- text to speech
- multiscale
- endpoint detection
- speaker recognition
- broadcast news
- pattern recognition
- multi stream
- dialogue system
- estimation algorithm
- spontaneous speech
- lexical features
- data sets
- automatic speech recognition systems