Classification of clean and noisy bilingual movie audio for speech-to-speech translation corpora design.
Andreas TsiartasPrasanta Kumar GhoshPanayiotis G. GeorgiouShrikanth S. NarayananPublished in: ICASSP (2014)
Keyphrases
- audio visual
- acoustic signals
- speech music discrimination
- broadcast news
- text to speech
- audio stream
- speech recognition
- speaker identification
- pattern recognition
- speech signal
- noisy environments
- prosodic features
- language resources
- chinese english
- machine translation
- speech synthesis
- audio signals
- emotion recognition
- audio features
- cepstral features
- spectral features
- acoustic signal
- information retrieval
- feature extraction
- statistical machine translation
- query translation
- cross language information retrieval
- feature space
- classification accuracy
- machine learning
- visual speech
- multi stream
- text classification
- natural language processing
- spontaneous speech
- comparable corpora
- feature selection
- text mining
- visual information
- parallel corpus
- parallel corpora
- speaker recognition
- machine translation system