Audio / Youtube to Noun Word-Cloud

Generate wordcloud from extracted nouns from speech audio

This is a python3 program used for generating a word cloud of the most commonly used nouns and proper nouns said in a particular youtube video or audio file (containing spoken word). This program utilizes pocketsphinx for speech-to-text and nltk for word tokenizing.

To generate a wordcloud from a youtube video:

python3 audio_noun_wordcloud.py --url=<youtube_video_url>

To generate a wordcloud from an audio .wav file on disk:

python3 audio_noun_wordcloud.py --path=<path_to_wav>

Options

Use the --save flag to save the generated word-cloud as a .png image to a specified path:

python3 audio_noun_wordcloud.py --url=<youtube_video_url> --save path/

Dependencies

• SpeechRecognition
• pocketsphinx
• ffmpeg
• youtube_dl
• nltk
• matplotlib
• wordcloud
• pydub