Enhance Image Generation with jhorovitz/omini-schnell Cognitive Actions

23 Apr 2025
Enhance Image Generation with jhorovitz/omini-schnell Cognitive Actions

In today's digital landscape, the ability to generate high-quality images programmatically is a powerful tool for developers across various domains. The jhorovitz/omini-schnell API offers a unique set of Cognitive Actions that allow developers to harness pretrained Diffusion Transformer models for image generation tasks. These actions enable users to incorporate subject control into scenes without the need for extensive training, preserving subject identity while facilitating flexible pose or scene changes. This article will guide you through the key features of the Incorporate Subject Control in Scenes action and how to integrate it into your applications effectively.

Prerequisites

Before diving into the integration process, ensure you have the following:

  • An API key for accessing the Cognitive Actions platform.
  • Basic knowledge of making API calls and handling JSON data in your preferred programming language.
  • Familiarity with Python, as we will use it for our conceptual examples.

For authentication, you will typically pass your API key in the headers of your requests.

Cognitive Actions Overview

Incorporate Subject Control in Scenes

The Incorporate Subject Control in Scenes action lets you insert items into a scene using predefined models. This action is particularly beneficial for generating images that require specific subjects to be maintained while changing the context around them.

Input

The input for this action requires a JSON object with the following schema:

{
  "seed": "integer",
  "prompt": "string",
  "taskName": "string",
  "controlImage": "string",
  "outputFormat": "string",
  "guidanceScale": "number",
  "outputQuality": "integer",
  "numberOfOutputs": "integer",
  "numberOfInferenceSteps": "integer"
}

Example Input:

{
  "prompt": "A diecast toy car of this, it is photographed in a toy box with the word \"Waymo\", the box is on a shop shelf selling for $5.99",
  "taskName": "subject_1024",
  "controlImage": "https://replicate.delivery/xezq/Za0hrhIO1PboAlVUVyfnqfkOX4y0gh6olDHcngJC9ifNIsgoA/tmpt_61zuhm.jpg",
  "outputFormat": "webp",
  "guidanceScale": 3.5,
  "outputQuality": 80,
  "numberOfOutputs": 4,
  "numberOfInferenceSteps": 8
}
  • seed: Optional. A random seed for reproducible generation.
  • prompt: A textual description guiding image generation.
  • taskName: Specifies the task type (options: subject_512, subject_1024, default is subject_1024).
  • controlImage: URL of the control image to guide the generation process.
  • outputFormat: Desired format for the output image (options: webp, jpg, png, default is webp).
  • guidanceScale: A scale factor for image guidance (between 0 and 10, default is 3.5).
  • outputQuality: Quality of the output images (between 0 and 100, default is 80, not relevant for .png).
  • numberOfOutputs: Number of image outputs desired (between 1 and 4, default is 1).
  • numberOfInferenceSteps: Total steps for the inference process (between 1 and 50, default is 8).

Output

The action typically returns an array of URLs pointing to the generated image outputs.

Example Output:

[
  "https://assets.cognitiveactions.com/invocations/fc9d3ce4-32d5-4f74-be99-04b9c324613c/f040e6c6-258c-4f5a-a189-f60ab31bb1df.webp",
  "https://assets.cognitiveactions.com/invocations/fc9d3ce4-32d5-4f74-be99-04b9c324613c/2e89b0af-440b-4f35-a9f4-4ee6b930bf4d.webp",
  "https://assets.cognitiveactions.com/invocations/fc9d3ce4-32d5-4f74-be99-04b9c324613c/419f25e7-5f77-4637-9113-a6182c290b74.webp",
  "https://assets.cognitiveactions.com/invocations/fc9d3ce4-32d5-4f74-be99-04b9c324613c/f6456698-7d50-4912-a8ef-4c6d3906b322.webp"
]

Conceptual Usage Example (Python)

Here’s how you might structure the API call in Python:

import requests
import json

# Replace with your Cognitive Actions API key and endpoint
COGNITIVE_ACTIONS_API_KEY = "YOUR_COGNITIVE_ACTIONS_API_KEY"
COGNITIVE_ACTIONS_EXECUTE_URL = "https://api.cognitiveactions.com/actions/execute" # Hypothetical endpoint

action_id = "82730be4-a579-4eb5-b82a-a7ff8753b5d2" # Action ID for Incorporate Subject Control in Scenes

# Construct the input payload based on the action's requirements
payload = {
    "prompt": "A diecast toy car of this, it is photographed in a toy box with the word \"Waymo\", the box is on a shop shelf selling for $5.99",
    "taskName": "subject_1024",
    "controlImage": "https://replicate.delivery/xezq/Za0hrhIO1PboAlVUVyfnqfkOX4y0gh6olDHcngJC9ifNIsgoA/tmpt_61zuhm.jpg",
    "outputFormat": "webp",
    "guidanceScale": 3.5,
    "outputQuality": 80,
    "numberOfOutputs": 4,
    "numberOfInferenceSteps": 8
}

headers = {
    "Authorization": f"Bearer {COGNITIVE_ACTIONS_API_KEY}",
    "Content-Type": "application/json"
}

try:
    response = requests.post(
        COGNITIVE_ACTIONS_EXECUTE_URL,
        headers=headers,
        json={"action_id": action_id, "inputs": payload} # Hypothetical structure
    )
    response.raise_for_status() # Raise an exception for bad status codes (4xx or 5xx)

    result = response.json()
    print("Action executed successfully:")
    print(json.dumps(result, indent=2))

except requests.exceptions.RequestException as e:
    print(f"Error executing action {action_id}: {e}")
    if e.response is not None:
        print(f"Response status: {e.response.status_code}")
        try:
            print(f"Response body: {e.response.json()}")
        except json.JSONDecodeError:
            print(f"Response body: {e.response.text}")

In this Python snippet:

  • Replace YOUR_COGNITIVE_ACTIONS_API_KEY with your actual API key.
  • The action_id corresponds to the Incorporate Subject Control in Scenes action.
  • The payload is structured to meet the input requirements, ensuring that all necessary fields are included.

Conclusion

The jhorovitz/omini-schnell Cognitive Actions provide powerful tools for developers looking to enhance their applications with advanced image generation capabilities. By integrating the Incorporate Subject Control in Scenes action, you can create dynamic and contextually rich images with minimal effort. Explore the possibilities of these actions in your projects, and consider how they can streamline your workflow or enhance user experiences. Happy coding!

← Back to All Articles