Fractal dimension feature for distinguishing between overlapped speech and single-speaker speech.
Wei LiQian-Hua HeYan-Xiong LiXueyuan ZhangXiao-Hvi FengPublished in: ICMLC (2012)
Keyphrases
- fractal dimension
- speech recognition
- audio visual
- automatic speech recognition
- speaker recognition
- speech signal
- speaker identification
- vocal tract
- speaker verification
- automatic speech recognition systems
- texture analysis
- speaker dependent
- prosodic features
- fractal analysis
- audio stream
- broadcast news
- text to speech
- speech synthesis
- speaker diarization
- neural network
- image processing
- pattern recognition
- fractal dimensions
- speech recognizer
- spontaneous speech
- feature extraction
- box counting
- speaker adaptation
- hidden markov models
- spoken language
- gaussian mixture model