Automatic Speech Recognition 2021/2022 Labs Code

This repository contains the code for the labs of Automatic Speech Recognition course in academic year 2021/22.

Here are some explanations about the files in this repo.

• asr_lab1.ipynb is a notebook containing the exercises for Lab 1

• asr_lab1_solutions.ipynb is a notebook containing the solutions for Lab 1

• introduction.pdf gives a refresher on representing HMMs in WFST form (covered in Lecture 5)

• helper_functions.py contains two functions, parse_lexicon() and generate_symbol_tables()

• lexicon.txt is a lexicon used in all labs

• lab1_template.py is the submission template for lab1.

• asr_lab2.ipynb is a notebook containing the exercises for Lab 2

• asr_lab2_solutions.ipynb is a notebook containing the solutions for Lab 2

• asr_lab3_4.ipynb is a notebook containing the exercises for Lab 3 and Lab4.

• asr_lab3_solutions.ipynb is a notebook containing the solutions for Lab 3

• asr_lab3_4_solutions.ipynb is a notebook containing the solutions for Lab 3 and Lab 4

• observation_model.py is the python file containing the observation_model that you need in Lab3.

• phonelist.txt is the phone list in our observation_model.py. For instance, there is a phone aa in phonelist.txt, so "aa_1" is a valid hmm_label when calling my_om.log_observation_probability(hmm_label, t).

• asr_lab5.ipynb is a notebook containing the exercises for Lab 5

• asr_lab5_solutions.ipynb is a notebook containing the solutions for Lab 5

For setup instructions, view the Piazza post