Highly accurate non-intrusive speech forensics for codec identifications from observed decoded signals.
Frank JennerAndres KwasinskiPublished in: ICASSP (2012)
Keyphrases
- highly accurate
- bitstream
- acoustic signals
- capable of producing
- high accuracy
- speech recognition
- signal processing
- fundamental frequency
- audio signals
- high quality
- bit rate
- coding scheme
- speaker identification
- distributed source coding
- noisy environments
- audio visual
- spoken language
- accurate models
- video quality
- video coding
- distributed video coding
- hidden markov models
- acoustic signal
- multiresolution