Skip to content

paion-data/tiger

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

title emoji colorFrom colorTo sdk sdk_version app_file pinned license short_description
Tiger
🎥
blue
indigo
streamlit
1.39.0
app.py
false
apache-2.0
Movie transcription for language learners

Hugging Face space badge Hugging Face sync status badge Apache License Badge

Tiger

Tiger is an automatic speech recognition application that takes a video file as input and generates a video with subtitles as well as downloadable .txt, .vtt, .srt transcription files. Tiger generates transcriptions using the OpenAI Whisper models.

The app is available on 🤗 Hugging Face space. Please check it out.

How Project Tiger Started

I don't believe NLP is inherently AI. Project Tiger, however, enlightened me on what way should AI makes NLP more powerful for people.

Paraphrasing movie lines has been my favorite approach to learn a foreign language. This was why I was able to speak close-native English in early days. I used this same tactic for German today as well.

One morning I was studying German with the famous German WW2 movie Generation War. Managing to find its subtitles123 were easy for me. A problem, however, soon arose as I proceeded with them: the subtitles had errors - missing couple of phrases every few other lines. This was a big NO for language learners like me. In old days 10 years ago, I would've reached a dead end...

Almost immediately, however, I had this idea of having AI figure out those missing words for me; so I tried OpenAI's Whisper which, quite surprisingly, yield 100% accurate results. That was the first occasion where AI produced significant solid values for my life as a language learner.

Being so excited, I decided to industrialize this process and hereby present it as Project Tiger on Hugging Face, an AI movie transcription service that targets specifically for language learners

Note

On name "Tiger"

The project name tiger was taken from the German heavy tank of World War II - Tiger I.

I chose this name because the project idea was spawned from the study of German language. More importantly, Tiger I has been widely recognized as a revolutionary tank being a crown jewels in the history of warfare technology:

"It is tempting to view the appearance of the Tiger tank as revolutionary, ... the end result was a machine that represented a quantum leap forward in tank design and changed everything - forever"4

As Tiger had steered the direction of next-generation heavy tank, this project has also revolutionized my view toward AI subconsciously and serves as an empirical guide on how I should make NLP more powerful for people with AI.

License

The use and distribution terms for tiger are covered by the Apache License, Version 2.0.

Footnotes

  1. https://www.opensubtitles.org/en/subtitles/6243166/generation-war-eine-andere-zeit-de ↩

  2. https://www.opensubtitles.org/en/subtitles/6243167/generation-war-ein-anderer-krieg-de ↩

  3. https://www.opensubtitles.org/en/subtitles/6243168/generation-war-ein-anderes-land-de ↩

  4. Tiger tank : Panzerkampfwagen VI Tiger I Ausf. E (SdKfz 181): owner's workshop manual, The Tank Museum, 2011. p.13 ↩

About

🤗 Empowering multilinguals to study language with movie

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 95.3%
  • Ruby 4.7%