Using Zoom Meeting Linux SDK to Get Transcription

Hey @chunsiong.zoom,

I am researching the possibilities of using the Zoom Meeting Linux SDK to get transcription from meetings. From my findings, the best approach I know so far is to call the API from an Electron app to api.zoom.us when the meeting ends to retrieve the transcription from the cloud recording.

Is this the most optimal way? My priority is to have a solution that does not incur any additional costs. Do you have any other ideas?

@hieudng1209 I’m not sure what you might define as additional cost.

The server needed to run the instance of Linux Meeting SDK could probably can be classified as additional cost, and as well as the development hours associated with the development and debugging.

The benefit (of using a meeting sdk linux bot) however would be to have access to real time transcription when they happen.

The other alternative which you have mentioned works best if you already have all the necessary plans in place.

Thank @chunsiong.zoom for your prompt response.

I am currently seeking any way to obtain transcription (speech-to-text) from the Meeting-Linux-SDK.

I have tried searching and reviewed your discussions, where you stated that Meeting-Linux-SDK cannot obtain transcription (only available on macOS and Windows) and there are no plans to add this feature.

But if at this time, this Linux SDK can obtain transcription, please guide me on how to obtain it.

Currently, I can only get raw video and raw audio, and I need transcription to create a customer requested bot.

@hieudng1209 Here are a couple of ways to go about doing this.

  1. Run the Linux Meeting SDK with your own speech to text service + raw audio.
  2. Run the Windows Meeting SDK (Possibly on a docker environment) to get the transcription via the meeting_closedcaption_interface object
  3. Get the transcription from the API endpoint
1 Like

Thank @chunsiong.zoom for your quick response.

I find the solution using Windows Meeting SDK (with Docker) very feasible to address this issue. However, our client requests building an electron-react application. Therefore, I’m currently using GitHub - zoom/meetingsdk-linux-raw-recording-sample to create a bot using a Linux container.

If I switch to using Windows Meeting SDK (with Docker) now, do you have any sample (or repository) that I can refer to?

I’m not sure what you might be refering to here.

Anything which runs on docker will be without GUI for user to interact with.

I don’t have a sample which exactly does what you want, but I have them as 2 separate projects. You will need to combine them yourself.

Disclaimer: While Meeting SDK Windows can run on Docker, there is a high chance it might stop working in future versions as there is high dependencies on Microsoft’s Windows Images support GUI apps (without user interaction) in docker environment

Running Meeting SDK on Docker

Running Meeting SDK to get Transcription

1 Like