https://github.com/seanoshea/voice_to_image

A proof-of-concept of voice to image using VertexAI
https://github.com/seanoshea/voice_to_image

Last synced: 22 days ago
JSON representation

A proof-of-concept of voice to image using VertexAI

• Host: GitHub
• URL: https://github.com/seanoshea/voice_to_image
• Owner: seanoshea
• License: other
• Created: 2024-06-14T21:22:07.000Z (7 months ago)
• Default Branch: feature/initial
• Last Pushed: 2024-09-12T02:36:44.000Z (4 months ago)
• Last Synced: 2024-10-25T08:35:16.291Z (2 months ago)
• Language: JavaScript
• Size: 85.9 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Voice to Image

A proof of concept node application which utilizes Vertex AI services to convert voice to images.

# Attribution

The overwhelming majority of the voice to text code (see `src/voice_listener.mjs`) was taken from https://cloud.google.com/speech-to-text/docs/transcribe-streaming-audio.

# Prerequistes

`npm run start` to get things going.

# key.json

You'll need to generate a `key.json` file to authenticate with VertexAI APIs. https://cloud.google.com/iam/docs/create-short-lived-credentials-direct for details. For reference, a `key.json` file should look a little like:

```

{

  "account": "",

  "client_id": "UUIDhur.apps.googleusercontent.com",

  "client_secret": "UUID",

  "quota_project_id": "voice-to-image-422913", // you will need to create a new project in the console and associate a billing account with it

  "refresh_token": "UUID",

  "type": "authorized_user",

  "universe_domain": "googleapis.com"

}

```

See https://stackoverflow.com/questions/70119129/refresh-token-gcp for some more details as well.

# Contributing

Suggestions and bug reports for the application are always welcome. Open an issue on github if you'd like to see an addition to the application or if you spot a bug.