Court

AI-based, unbiased, and transparent video analysis

Installation

Clone the repository:

git clone https://github.com/lilylilylily123/court.git

Make sure you have a mp4 file ready for analysis, preferably in the same directory as Court.

IMPORTANT

ffmpeg Installation

Ensure that you have ffmpeg installed on your system. You can download it from here.

Whisper Installation

Clone the repository:

git clone https://github.com/ggerganov/whisper.cpp

Build the project:

cd whisper.cpp
bash models/download-ggml-model.sh base
make

Your whisper executable path should be:

./build/bin/whisper-cli

but we recommend using absolute paths.

.env setup

Follow .env.example

Usage

python3 video_analyzer.py your_video.mp4

Output

A file named summary.txt will contain the summary of the video.

For more details, you can check the transcript.txt file, which contains the raw Whisper transcription.

Warning:

This project is not token-friendly, and can consume quite a lot of tokens, especially when analyzing longer videos. Ensure that you have enough tokens available before analyzing a video.

Credits

Whisper: https://github.com/ggerganov/whisper.cpp