Improving automatic speech recognition in noise by energy normalization and signal resynthesis.
Mircea GiurgiuAhsanul KabirPublished in: ICCP (2011)
Keyphrases
- automatic speech recognition
- noisy environments
- spectral subtraction
- speech recognition
- speech signal
- random noise
- speech segments
- hidden markov models
- noise model
- speech enhancement
- word error rate
- noise reduction
- broadcast news
- speech retrieval
- conversational speech
- bandpass
- low signal to noise ratio
- spoken words
- speech corpus
- noise level
- non stationary
- pattern recognition
- recognition errors
- spontaneous speech
- power spectrum
- word recognition
- machine learning
- noisy images
- signal to noise ratio
- denoising
- feature extraction
- information retrieval