Spaces:
Sleeping
Sleeping
| title: Smart Edit Assistant | |
| emoji: 🎬 | |
| colorFrom: blue | |
| colorTo: indigo | |
| sdk: streamlit | |
| app_file: app.py | |
| pinned: false | |
| hardware: gpu | |
| hf_oauth: true | |
| hf_oauth_scopes: | |
| sdk_version: 1.44.1 | |
| # Smart Edit Assistant | |
| **Smart Edit Assistant** is an AI-powered web application that **automates video editing tasks** end-to-end. Users can upload video files, let the system **extract audio**, **transcribe** the speech (e.g., via Whisper), **analyze** content with GPT-like models, and **apply automated cuts and edits** using FFmpeg or MoviePy. The end result is a curated, shorter (or otherwise improved) video that can be downloaded, saving creators time on manual post-production. | |
| ## Features | |
| - **Video Upload & Preview**: Upload `.mp4`, `.mov`, or `.mkv` files. | |
| - **Audio Extraction**: Efficiently pulls the audio track for transcription. | |
| - **AI Transcription**: Uses OpenAI Whisper (API or local) or other STT solutions. | |
| - **LLM Content Analysis**: GPT-4 or open-source LLM suggests cuts and highlight segments. | |
| - **Automated Editing**: Uses FFmpeg/MoviePy to cut and reassemble segments, optionally insert transitions. | |
| - **Result Preview**: Plays the edited video in-browser before download. | |
| - **(Optional) User Authentication**: Configurable free vs. premium tiers. | |
| ## Repository Structure | |
| smart-edit-assistant/ ├── app.py # Main Streamlit app ├── pipelines/ │ ├── video_process.py # Audio extraction & editing logic (MoviePy / FFmpeg) │ ├── ai_inference.py # Whisper/GPT calls for transcription & instructions │ └── auth_utils.py # Optional authentication logic ├── .streamlit/ │ └── config.toml # Streamlit config (upload limit, theming) ├── requirements.txt # Python dependencies ├── apt.txt # (Optional) System-level dependencies if needed └── README.md # Project description (this file) | |
| bash | |
| Copy code | |
| ## Local Development & Setup | |
| 1. **Clone this repo**: | |
| ```bash | |
| git clone https://github.com/YourUsername/smart-edit-assistant.git | |
| cd smart-edit-assistant | |
| Install Python dependencies: | |
| bash | |
| Copy code | |
| pip install -r requirements.txt | |
| If you plan to run open-source Whisper locally, ensure you install openai-whisper or the GitHub repo (git+https://github.com/openai/whisper.git). | |
| If you’re using GPU, make sure your PyTorch install matches your CUDA version. | |
| Run the app: | |
| bash | |
| Copy code | |
| streamlit run app.py | |
| Open http://localhost:8501 in your browser to interact with the UI. | |
| Set Environment Variables (for GPT or Whisper API, if needed): | |
| bash | |
| Copy code | |
| export OPENAI_API_KEY="sk-..." | |
| or store in a local .env file and load with python-dotenv. | |
| Deploying on Hugging Face Spaces | |
| Create a Space: | |
| Go to Hugging Face Spaces and create a new Space with the Streamlit SDK option. | |
| Upload your files: | |
| Either drag-and-drop via the web interface or push via Git: | |
| bash | |
| Copy code | |
| git remote add origin https://huggingface.co/spaces/YourUsername/Smart-Edit-Assistant | |
| git push origin main | |
| Set your secrets: | |
| In the Space’s Settings page, add OPENAI_API_KEY or any other API keys under “Secrets”. | |
| If you want GPU, set hardware: "gpu" in the YAML frontmatter (as shown above) or in the Space settings. | |
| Build and Launch: | |
| The Space will automatically install your requirements.txt and run app.py. | |
| Once deployed, your app is live at https://huggingface.co/spaces/YourUsername/Smart-Edit-Assistant. | |
| Usage | |
| Upload a Video: Click “Browse files” to select a .mp4, .mov, or .mkv file. | |
| Extract & Transcribe: The app automatically pulls the audio, then uses Whisper or another STT method to get a transcript. | |
| Generate Edits: An LLM (GPT-4 or local) analyzes the transcript and suggests where to cut or remove filler content. | |
| Apply Edits: The app runs ffmpeg or MoviePy to create a new edited video file. | |
| Preview & Download: You can watch the edited clip directly in the browser and then download the .mp4. | |
| Configuration | |
| Streamlit Config: | |
| .streamlit/config.toml can set maxUploadSize (e.g. 10GB) or color theme. | |
| Authentication: | |
| If hf_oauth is true, users must log in with their Hugging Face account. | |
| For custom username/password or free vs. premium tiers, see auth_utils.py or documentation in your code. | |
| Roadmap | |
| Interactive Timeline: Let users manually tweak the AI’s suggested cuts. | |
| B-roll Insertion: Generate or fetch recommended B-roll and splice it in automatically. | |
| Transition Effects: Provide crossfades, text overlays, or AI-generated intros/outros. | |
| Multi-user Collaboration: Shared editing session or project saving in a database. | |
| Troubleshooting | |
| File Not Found or Zero Bytes: Make sure ffmpeg or MoviePy didn’t fail silently. Check logs for errors. | |
| Whisper “load_model” Error: Ensure you installed openai-whisper or the GitHub repo, not the unrelated “whisper” PyPI package. | |
| Large File Upload: If large uploads fail, confirm the maxUploadSize in .streamlit/config.toml is high enough, and verify huggingface secrets/config. | |
| Performance: For best speed, request a GPU from Hugging Face Spaces or use a local GPU with the correct PyTorch/CUDA version. | |
| License | |
| You can choose a license that suits your project. For example: | |
| java | |
| Copy code | |
| MIT License | |
| Copyright (c) 2025 ... | |
| Permission is hereby granted, free of charge, to any person obtaining a copy | |
| of this software and associated documentation files (the "Software"), ... |