SpeakerLDA: Discovering Topics in Transcribed Multi-Speaker Audio Contents.
Damiano SpinaJohanne R. TrippasLawrence CavedonMark SandersonPublished in: SLAM@ACM Multimedia (2015)
Keyphrases
- audio visual
- multimedia
- spontaneous speech
- automatic transcription
- speaker identification
- prosodic features
- multi modal
- visual information
- signal processing
- topic models
- speech recognition
- related topics
- audio stream
- acoustic features
- visual data
- information retrieval
- database
- latent dirichlet allocation
- gaussian mixture model
- web content
- hidden markov models
- feature extraction
- metadata
- search engine
- data mining