AWS Artificial Intelligence Hackathon 2018

End to end Video Translation Platform

Imagine watching any youtube video (eg: Documentary) with audio in your own language. We built an end to end video translation platform using AWS AI services that just does that in one click without human intervention.

Uniqueness

Enables the power to watch any video in the language of choice of the user. Imagine watching a course video from Udacity/Coursera in the native language. Imagine learning to cook a Chinese dish by translating a video in native Chinese to English. Imagine dubbing a video automatically into multiple languages with a single click.

I have tested it on even documentaries of 20-30 minutes long and it works very well to synthesize at the appropriate timecodes without any offset.

Input:

The input to the program is a youtube video url link in any language. Works best for single speaker videos (Eg: Documentaries, Monologues etc). Currently only English and Spanish are supported by Amazon Transcribe. So input youtube url link should be in English(US) or Spanish(US).

Output

The output is a downloaded video with the audio track in the language of your choice.

A demo can be seen at: https://www.youtube.com/playlist?list=PLeIs-OgljgIyrYeAp7K26I88KfOePJHu4

In the above demo you can see a playlist with two videos.
1st Video : English Original Audio
2nd Video : Translated Audio in Chinese (English subtitles available generated using Amazon Transcribe)

Installation

The installation instructions are available in the readme of the GitHub repository: https://github.com/ramsrigouthamg/aws_artificial_intelligence_hackathon_2018

"Devposttesting" has been added to the repository. Here is the invite link again: https://github.com/ramsrigouthamg/aws_artificial_intelligence_hackathon_2018/invitations

Algorithm

Services used: Amazon Transcribe, Amazon Translate, Amazon Polly, Lambda (3 different S3 lambda triggers). The flowchart provided in the image gallery explains the solution. Detailed instructions are available in the Readme.md file in the project github repository.

Built With

  • amazon-lambda
  • amazon-polly
  • amazon-transcribe
  • amazon-translate
+ 3 more
Share this project:
×

Updates