Added if __name__ == "__main__" check.

Change WelshAudioProcessor to PhonemeLevelSpeechProcessor for clearer representation of it's purpose and utility. - It can be used for any phonetic language

- Added check for dictionary existing
- Added function for loading corpus for clarity
- Added check if corpus file exists in load function

- Refactoring

- Check if the entered word Is numeric before trying to convert to int()

These will be used in the separate testing branch instead. Removed to make the main branch's purpose clearer.

- Function called in init, and can now be updated directly from the menu, for rapid on the fly changes and testing of additions to the dictionary file.

Added variables for the respective language dictionary and corpus files

- Updated temp files

- Removed 'dictionary.txt' and 'corpus.txt' from directory
- Added spanish and welsh equivalents for both dictionary and corpus

Contains Installation instructions

3 seconds reduced to 2 seconds

seect -> select

Changed spelling to correct version.

Class can now provide audio output for words in the corpus, as long as the file exists.

Added [l] option to menu, where the selected word will be played

A start and stop sound is now played on recording start and stop

Added recordings of correct pronunciations for words in corpus

Contains python package requirements.txt for use in setting up environment for syste use.

Utilises interactions class to perform all of the functionality of the system in a readable format

Class contains the reorder, processor and checker classes, and can be utilised by main.py to supply user with menus, record, process and compare input speech.

__init__ - loads the classes mentioned above. Adds a local variable for the mode and exit

modeSelector - shows the mode selector menu and allows mode selection

wordSelector - shows the word selection menu and allows word selection

repeatMenu - shows a menu after one round of use/recording, which allows user to exit, change mode or word, or to record again

printCurrentWord - prints the currently selected word which is being trained on

check - checks the transcription against the selected word, returning the appropriate output

recordAndProcess - records the audio using the reorder class and processes it using the processor class, outputting the model's transcription

__init__ - gets the corpus from the corpus file, sets a variable ready to store the currently selected word.

getCorpus - returns the corpus as a list of str

setWord - sets the local word variable

compare - Compares a supplied string against the currently selected word, and formats output accordingly. Shows a list of differences across the two words if there are any.

Added class to deal with recording

__init__ - Sets sample rate, recording chunk size, time to record for, and creates a place in memory to store audio frames.

record - records the audio using the previously set parameters, returning a numpy.array for processing

List of 15 words which are known to work with the system

Contains Methods:
__init__ - Loads the wav2vec2 model, gets the dictionary from dictionary.txt in root dir

processWavDir - redacted to reduce footprint, can be uncommented for use in testing.

processAudio - processes a numpy array and transcribes using International Phonetic Alphabet (IPA)

convertIPAToRoman - converts the IPA word into corresponding welsh orthography (latin script) using the disctionary