The pace at which content creation is in progress is simply outstanding. I'm especially interested in talking about videos, as today's article is about language translation for YouTube videos. It can help people access information and resources otherwise not consumable due to language barriers. YouTube offers myriad content, including music videos, educational content, comedy skits, and more, catering to a wide range of interests. The platform's popularity is at a different level due to increased global internet penetration.
I post my learnings on my channel dedicated to AWS cloud technology AWSLearn. The YouTube channel is in English, but I was thinking, can I make videos available in more languages? The subtitles are helpful but not as much compared to the audio language. My curiosity kept me thinking; I came across the AWS services catering to a similar use case. I decided to try them. I know the Hindi language and gave it a try. So English to Hindi video translation.
Let's see the first sample of my work; we will also go into more detail.
English Video :
Hindi Video:
The translation accuracy is not 100%, however, I think there are more options in these services to customize and improve accuracy. I can give it a shot sometime in the future. The idea was only to explore how to achieve translation. You can choose from the list of languages for the translation and Hindi here is just an example.
The workflow is as follows:
- Video file is uploaded to S3 bucket.
- It triggers lambda function; Amazon Transcribe converts the speech to text.
- Amazon Translate converts the language text from English to Hindi.
- Amazon Polly converts text to speech in Hindi.
- Audio is outputted to S3 bucket or even can be downloaded to desktop/laptop.
- Use any video editing software to replace audio file.
Understand more about translation services from Amazon.
Amazon Transcribe
Amazon Transcribe is an AWS Artificial Intelligence (AI) service that makes it easy for you to convert speech to text. Using Automatic Speech Recognition (ASR) technology, you can use Amazon Transcribe for a variety of business applications, including transcription of voice-based customer service calls, generation of subtitles on audio/video content, and conduct (text-based) content analysis on audio/video content.Amazon Translate
Amazon Translate is a Neural Machine Translation (MT) service for translating text between supported languages. Powered by deep learning methods, the service provides high-quality, affordable, and customizable language translation, enabling developers to translate company and user-authored content, or build applications requiring support across multiple languages. The service can be used via an API, enabling either real-time or batch translation of text from the source language to the target language.Amazon Polly
Amazon Polly is a service that turns text into lifelike speech. Amazon Polly enables existing applications to speak as a first class feature and creates the opportunity for entirely new categories of speech-enabled products, from mobile apps and cars, to devices and appliances. Amazon Polly’s pay-as-you-go pricing, low cost per request, and lack of restrictions on storage and reuse of voice output make it a cost-effective way to enable speech synthesis everywhere.
This is a conceptual project at this stage and only part of it is automated using Lambda functions written in Python. I will have more details coming to this article. Hopefully the solution presented in this article will trigger more thoughts and we will witness exciting new architectures and amazing projects.
Keep figuring out !!
Top comments (0)