VoiceFlow Add-on

This Adobe Express Add-on allows you to translate text within images using the "Nano Banana" (Google Gemini) API.

Features

• Automatic Workflow: Select an image on the canvas, translate it, and the result is automatically added back to the document.
• Image Upload: Alternatively, upload an image file directly.
• Custom Prompts: Choose a target language or write a custom prompt for the AI.

Setup

1. Backend Setup

The backend handles the communication with the Google Gemini API.

1. Navigate to the backend directory:

   cd backend

2. Install dependencies (if not already done):

   npm install

3. Important: Open backend/.env and replace YOUR_API_KEY_HERE with your actual Google Gemini API key.
4. Start the server:

   npm start

   The server will run on http://localhost:3000.

2. Frontend (Add-on) Setup

1. Navigate to the hello-world directory:

   cd hello-world

2. Install dependencies:

   npm install

3. Start the add-on:

   npm start

4. Follow the instructions in the terminal to open the add-on in Adobe Express.

Usage

1. Select an Image: In Adobe Express, select the image you want to translate on the canvas.
2. Get from Canvas: Click "Get from Canvas" in the add-on panel.
3. Translate: Choose a language and click "Translate Image".
4. Result: The translated image will be displayed and automatically added to your canvas.

Notes

• The translation relies on the gemini-3-image-preview model.
• "Get from Canvas" relies on the Add-on SDK's rendition capabilities. If it fails for a specific element type, try taking a screenshot or using the "Upload File" option.