A tutorial notebook with examples with scalar data, multidimensional data and real speech data.
It also includes examples of warping identical sentences spoken by different speakers such that they can be combined in sync with each other as you would do for lipsyncing.
notebook: Levenshtein_ex.ipynb
- Exercise 1: Computing the Levenshtein Distance
- do the exercise for one of the suggested word pairs
- answer the additional question on the difference between case 2 & 3
- Exercise 2: character vs. word Levenshtein Distance
- Exercise 3: Error rate computations for speech recognition
notebook: DTW_ex.ipynb
- Exercise 1: DTW on 1-dimensional feature vector sequence
- Exercise 2: Using different warping constraints
... coming soon ...