Clips - Open Source AI Clipping Tool

What is Clips AI?

Clips AI is an open-source Python library designed to simplify the process of converting long-form videos into shorter, more manageable clips. It is particularly useful for audio-centric, narrative-based videos such as podcasts, interviews, speeches, and sermons. With Clips AI, users can automatically segment videos into multiple clips and resize the aspect ratio from 16:9 to 9:16, making it easier to share and view on different platforms.

Benefits

Clips AI offers several key advantages:

• Automatic Clipping: The tool automatically identifies and creates clips from long-form videos, saving time and effort.
• Dynamic Resizing: It dynamically reframes videos to focus on the current speaker, ensuring that the content is presented in the most engaging way possible.
• Open-Source: Being open-source, Clips AI is free to use and can be customized to meet specific needs.
• Easy Integration: With just a few lines of code, users can integrate Clips AI into their existing workflows.

Use Cases

Clips AI is ideal for a variety of applications, including:

• Podcast Production: Automatically segmenting podcast episodes into shorter clips for social media sharing.
• Interview Transcription: Creating clips from interviews to highlight key moments and insights.
• Speech and Sermon Archiving: Breaking down long speeches or sermons into smaller, more digestible segments.
• Content Creation: Resizing videos to different aspect ratios for optimal viewing on various platforms.

Installation

To get started with Clips AI, users need to install Python dependencies. It is recommended to use a virtual environment to avoid dependency conflicts. The installation process involves running a few simple commands:

pip install clipsaipip install whisperx@git+https://github.com/m-bain/whisperx.git

Usage

Clips AI uses the video's transcript to identify and create clips. The process involves transcribing the video and then using the transcription to find clips. Here is a basic example of how to use Clips AI:

from clipsai import ClipFinder, Transcribertranscriber = Transcriber()transcription = transcriber.transcribe(audio_file_path="/abs/path/to/video.mp4")clipfinder = ClipFinder()clips = clipfinder.find_clips(transcription=transcription)print("StartTime: ", clips[0].start_time)print("EndTime: ", clips[0].end_time)

For resizing the video, users need a Hugging Face access token. The resizing process involves running a few more lines of code:

from clipsai import resizecrops = resize(video_file_path="/abs/path/to/video.mp4",pyannote_auth_token="pyannote_token",aspect_ratio=(9, 16))print("Crops: ", crops.segments)