Audio-based automatic detection of objectionable contents in noisy conditions using normalized segmental two-dimesional MFCC.
Bong-Wan KimDae-Lim ChoiJae-Deok LimSeungWan HanYong-Ju LeePublished in: ICASSP (2012)
Keyphrases
- automatic detection
- multimedia
- automatically detecting
- small bowel
- speaker identification
- sufficient conditions
- hidden markov models
- noisy environments
- automatic classification
- audio visual
- linear predictive
- acoustic features
- gaussian mixture model
- speech recognition
- feature extraction
- metadata
- speech signal
- computer assisted
- audio features
- feature set
- mel frequency cepstral coefficients
- feature vectors
- similarity measure