Skip to content

Latest commit

 

History

History
30 lines (24 loc) · 1.83 KB

README.md

File metadata and controls

30 lines (24 loc) · 1.83 KB
Raw

Find and Analyze Videos using Azure OpenAI GPT4-Vision with Video Enhancements

drawing

This solution accelerator presents a detailed framework for analyzing videos, leveraging Azure OpenAI GPT4-Vision with Video Enhancements technology.
It's a crucial resource for businesses across a range of industries, including marketing, media, education, manufacturing, healthcare, and retail.

Getting started

  1. Create an Azure Storage Account resource
    • Create a Blob container named videos
    • Generate a SAS token for the videos container with Read, Add, Create, Write, Delete and List permissions. Ensure that the Expiry date covers the duration of your use case.
  2. Create an Azure OpenAI resource in the Sweden Central or Switzerland North region
    • Deploy the GPT4 vision-preview model
  3. Create an Azure Computer Vision resource in the same region as the OpenAI Service
  4. Use your preferred Python IDE to open the find-and-analyze-videos.ipynb notebook

    • We have developed it in a conda environment on GitHUb Codespaces with Python 3.10
      bash conda create -n vision python=3.10

    • Using VSCode ensures full compatibility with video playback in the notebook

    • Use the requirements.txt file to install the required packages in your environment
      bash pip install -r requirements.txt

  5. Rename the dotenv template to .env and populate it with the required parameters and credentials for the Azure resources.
  6. We have provided a couple of sample videos in the videos folder for a quick start. Feel free to add your own videos.

Note: In case of an import error related to libGL.so.1 and the cv2 library, you can manually install the required components (Ubuntu example):

sudo apt-get update
sudo apt-get install libgl1-mesa-glx