RViewer

RViewer is a VLC media player that can generate subtitle using OpenAI Whisper Model. It's a personal project that I developed because I wanted to create a video player that was comfortable to use for many hours a day for video lessons (or in general to analyze video). In particular, I used WhisperX.

sample.mp4

prerequisites:

• python3
• VLC program
• [Optional] (to generate subtitles): ffmpeg program

libraries:

• python-vlc, pyside6
• [Optional] (to generate subtitles): ffmpeg-python, whisperx, torch

hardware requirements (to generate subtitles):

• Windows/MacOS/Linux
• at least 2GB free space for libraries/models
• at least 1GB VRAM for OpenAI Tiny model, 3GB VRAM for Small model

setup

1. Open a terminal and clone the repo: gh repo clone Luruu/RViewer (or download it manually)
2. Move to RViewer folder: cd RViewer
3. Create and activate a virtual environment:
   • Linux/MacOS: python3 -m venv env and source env/bin/activate
   • Windows: py -m venv env and .\env\Scripts\activate
4. Install libraries: pip install pyside6 python-vlc git+https://github.com/m-bain/whisperx.git

note: you must have the VLC program installed to be able to view the videos (and open RViewer) and the ffmpeg program to be able to use whisper.

tips

• (Subtitle): use CUDA if you have a NVIDIA GPU: it will be faster. Check if torch.cuda is available with: test_torch.py. This helped me to activate it: pytorch/pytorch#30664 (comment)
• change content of CSS files if you want to change object styles
• change vlc_args.txt in preferences folder if you want to change the behavior of vlc (for example, to change subtitle font, size and so on)

solutions

• [MacOS]: if you get error SSL: CERTIFICATE_VERIFY_FAILED try to install certifi or upgrade it.

possible future updates

1. add code documentation
2. handle playlists
3. handle youtube videos
4. Generate audio transcript from subtitles
5. vocal commands
6. Stream video

credits

Program Icon created by Azland Studio - Flaticon

Outputs