Speech Collage: Code-Switched Audio Generation by Collaging Monolingual Corpora.
Amir HusseinDorsa ZeinaliOndrej KlejchMatthew WiesnerBrian YanShammur Absar ChowdhuryAhmed AliShinji WatanabeSanjeev KhudanpurPublished in: ICASSP (2024)
Keyphrases
- audio visual
- audio stream
- audio signals
- broadcast news
- speaker identification
- chinese english
- text to speech
- audio features
- digital audio
- parallel corpus
- emotion recognition
- speech processing
- question answering
- audio recordings
- natural language processing
- cepstral features
- speech music discrimination
- acoustic signals
- linear predictive coding
- statistical machine translation
- query expansion
- domain specific
- source code
- speech recognition
- cross language information retrieval
- cross lingual
- information retrieval
- multimedia
- information extraction
- audio files
- linguistic resources
- audio video
- text data
- visual speech
- multi modal
- prosodic features
- cross language
- speech signal
- ad hoc retrieval
- machine translation
- visual information
- audio signal
- spoken language
- automatic transcription
- automatic speech recognition