Spectro-temporal directional derivative features for automatic speech recognition.
James GibsonMaarten Van SegbroeckAntonio OrtegaPanayiotis G. GeorgiouShrikanth S. NarayananPublished in: INTERSPEECH (2013)
Keyphrases
- automatic speech recognition
- speech recognition
- feature extraction
- low level
- hidden markov models
- feature set
- feature space
- temporal information
- recognition errors
- information retrieval systems
- document collections
- non stationary
- speech signal
- image processing
- noisy environments
- word recognition
- word error rate
- conversational speech