Generative Remove

Remove unwanted objects from images using Cloudinary's AI-powered generative removal technology. This method intelligently removes specified objects and fills the area with contextually appropriate content.

Overview

The Generative Remove feature uses Cloudinary's advanced AI to:

• Smart Object Detection: Identifies and removes objects based on text prompts
• Contextual Fill: Generates realistic content to fill removed areas
• Shadow Removal: Optionally removes shadows cast by removed objects
• Multiple Instance Support: Can remove all instances of an object in one operation

Basic Usage

import { ImagineoAIClient } from '@imagineoai/javascript';

const client = new ImagineoAIClient(apiUrl, { 
  apiKey: process.env.IMAGINEOAI_API_KEY 
});

// Remove an object from an image
const result = await client.images.genRemove({
  image_url: "https://example.com/photo-with-unwanted-object.jpg",
  prompt: "person",
  remove_shadow: true,
  multiple: true
});

console.log("Object removed:", result);
// { 
//   url: "https://cdn.imagineoai.com/processed-image.jpg",
//   runId: "abc-123-def-456",
//   parentRunId: undefined
// }

Parameters

Parameter	Type	Required	Default	Description
image_url	string	One of image_url or run_id	-	URL of the source image
run_id	string	One of image_url or run_id	-	ID of a previous run to use as source
prompt	string	Yes	-	Text description of what to remove (e.g., "person", "car", "text")
remove_shadow	boolean	No	true	Whether to remove shadows cast by the removed object
multiple	boolean	No	true	Whether to remove all instances of the prompt

Examples

Remove a Single Object

// Remove a specific object from an image
const result = await client.images.genRemove({
  image_url: "https://example.com/street-photo.jpg",
  prompt: "trash can",
  remove_shadow: true,
  multiple: false  // Only remove one instance
});

Remove Multiple Objects

// Remove all people from a landscape photo
const result = await client.images.genRemove({
  image_url: "https://example.com/crowded-landscape.jpg",
  prompt: "person",
  remove_shadow: true,
  multiple: true  // Remove all people
});

Using Previous Run as Source

// Remove objects from a previously generated/processed image
const result = await client.images.genRemove({
  run_id: "previous-run-id-123",
  prompt: "watermark",
  remove_shadow: false,  // Watermarks don't cast shadows
  multiple: true
});

console.log("Watermarks removed from run:", result.parentRunId);

Complex Object Descriptions

// Remove specific objects with detailed descriptions
const result = await client.images.genRemove({
  image_url: "https://example.com/interior-photo.jpg",
  prompt: "red chair",  // Be specific about what to remove
  remove_shadow: true,
  multiple: false
});

Common Use Cases

E-commerce Product Photography

// Clean up product photos by removing unwanted elements
const result = await client.images.genRemove({
  image_url: productImageUrl,
  prompt: "price tag",
  remove_shadow: true,
  multiple: true
});

Real Estate Photography

// Remove temporary objects from property photos
const result = await client.images.genRemove({
  image_url: propertyPhotoUrl,
  prompt: "for sale sign",
  remove_shadow: true,
  multiple: false
});

Social Media Content

// Remove photobombers or unwanted people
const result = await client.images.genRemove({
  image_url: photoUrl,
  prompt: "person in background",
  remove_shadow: true,
  multiple: true
});

Response

The method returns a promise that resolves to:

{
  url: string;        // URL of the processed image
  runId: string;      // Unique ID for this operation
  parentRunId?: string; // ID of the source run (if using run_id input)
}

Best Practices

1. Be Specific with Prompts: More specific prompts yield better results

   • Good: "red car", "person wearing hat", "street lamp"
   • Less effective: "thing", "object", "stuff"

2. Shadow Removal: Enable remove_shadow when removing objects that cast visible shadows

3. Multiple Instances: Use multiple: true when you want to remove all instances of an object type

4. Image Quality: Higher resolution source images generally produce better results

5. Complex Scenes: For complex removals, consider multiple passes with different prompts

Limitations

• Works best with clearly defined objects
• May struggle with transparent or reflective objects
• Very large objects that dominate the image may leave visible artifacts
• Highly detailed backgrounds may show some inconsistencies after removal

Error Handling

try {
  const result = await client.images.genRemove({
    image_url: imageUrl,
    prompt: "unwanted object"
  });
  console.log("Success:", result);
} catch (error) {
  if (error.message.includes("credits")) {
    console.error("Insufficient credits");
  } else if (error.message.includes("not found")) {
    console.error("Run ID not found");
  } else {
    console.error("Generation failed:", error);
  }
}

Pricing

This operation consumes credits based on your Cloudinary plan. Check your usage limits before processing large batches of images.