Proof Of Concept based on Serverless generating subtitles from a video.
** Warning ** For this POC to work, we need a < 1minute-length video, as the Google Speech API Call we use only accepts videos of such length. For longer videos, we would need to use the long_running_recognize function, which is for technical reasons not working on Lambda, due to multithreading and Semaphores.
You will need to set up a Google Account and create a related project along with an AWS account.
In order to interact with Google Cloud Storage, Google Speech & Google Translate, you will need to create a Service Account Key.
Now, put it into the code/SpeechExtractor directory under a file named .credentials.json. Then, terraform apply the project and you should be all set!