Video Transcriptions

Demo of using Whisper for video transcriptions.

Setup

Follow set up instructions at https://github.com/openai/whisper.

• Install OpenAI Whisper:

pip install -U openai-whisper

• Install ffmpeg into your environment:

sudo apt update && sudo apt install ffmpeg

• Run using python3

python3 transcribe.py 

import whisper
import os

model = whisper.load_model("turbo")
for file in os.listdir("data"):
    if file.endswith((".mp4", ".mkv", ".mov", ".mp3", ".wav")):
        path = os.path.join("data", file)
        result = model.transcribe(path)
        with open(path + ".srt", "w") as f:
            f.write(result["text"])

Demo

Transcripts were generated for two samples:

• Free Surgical Care in Sudan: An Interview With Dr. Gina Portella - Whisper Transcript file.
• Safe Cerebral Protection: How to Handle Dangerous Sternal Reentry - Whisper Transcript file.