reference ] .right[ test utterance
] --- # DTW with Feature Data - Raw sample data way to numerous - Compute spectral (cepstral) features, eg. [MFCC](https://en.wikipedia.org/wiki/Mel-frequency_cepstrum) - Use euclidean distance on feature vectors - Uniform cost for substitution/deletion/insertion ## Isolated Word Recognition - Have prototype recordings for each word - Compute "distance" between test recoding and prototype words - Choose for word with least difference --- # Decoding States with DP ![dtmf1](/sequence-learning/02-cost-and-states/dtmf1.png) --- # Decoding States with DP ![dtmf2](/sequence-learning/02-cost-and-states/dtmf2.png) --- # Decoding States with DP - Don't compare to _individual class prototypes_ - Compare to _class prototypes_ - Since we relaxed the time constraint on the prototype axis, you must compute the minimum for all state transitions.