Add speech-to-text alignment, QC, word-to-TR alignment scripts + environment set up files

[complexbrains]

This PR adds the scripts and Python package versions to extract the language annotations for the CneuroMod naturalistic stimuli sets.

Codes were developed by @MarieStLaurent as part of the Algonouts project for the Friends dataset and refactored and adapted to be used for the rest of the @courtois-neuromod naturalistic datasets.

• speech-to-text_AA.py: Extracts the transcripts from .mkv files of the movies, aligns them with the audio and extracts time stamps of each word using AssemblyAI
• caption_timing_QC.ipynb: Takes word time stamps extracted by speech-to-text_AA.py and creates karaoke-style captions that play simultaneously over audio to check the quality of the alignment**
• text_align_AA.py : Finds the words feel into TR (fMRI repetition time) windows, and creates word chunks per each volume.
• requirement_caption_timing_QC.txt: Includes pip freeze of the venv used for the QC virtual environment.
• requirement_speech_to_text_TR_aligment.txt: Includes pip freeze of the venv used for the QC virtual environment.

[complexbrains]

@bpinsard @MarieStLaurent I cannot add reviewers, unfortunately! So pinging here 😊