Unlocking Image Class Detection with CudaNexus Detic Cognitive Actions

22 Apr 2025
Unlocking Image Class Detection with CudaNexus Detic Cognitive Actions

In today’s visually driven digital landscape, the ability to analyze and categorize images effectively is crucial for developers looking to enhance user experiences. The CudaNexus Detic offers a powerful solution through its Cognitive Actions, which leverage advanced techniques to detect a wide range of image classes. With the capability to identify up to twenty-thousand image classes using CLIP and ImageNet-21K, these actions provide developers with pre-built functionalities that can be integrated into their applications seamlessly.

Prerequisites

To start using the CudaNexus Detic Cognitive Actions, make sure you have the following:

  • An API key for the Cognitive Actions platform, which you will use for authentication.
  • Basic familiarity with making HTTP requests and handling JSON data.

Authentication usually involves passing your API key in the headers of your requests, ensuring secure access to the Cognitive Actions.

Cognitive Actions Overview

Detect Classes Using Detic

The Detect Classes Using Detic action enables you to detect various classes in images, offering a high degree of accuracy and flexibility in class selection.

Purpose

This action allows you to analyze images and identify objects within them using sophisticated models, making it ideal for applications involving image recognition, content moderation, and more.

Input

The input schema for this action requires the following fields:

  • image (required): A valid URI pointing to an image file that you want to analyze.
  • vocabulary (optional): The vocabulary set to be used for class detection. It defaults to "lvis," but you can choose from "lvis," "objects365," "openimages," "coco," or "custom."
  • customVocabulary (optional): A comma-separated string of custom vocabularies, applicable only when the vocabulary is set to "custom."

Example Input:

{
  "image": "https://replicate.delivery/pbxt/K6zgH3un6xi4l07DFSUOgHKzOIbWktlvmc348VZbRDqSbqeV/k67kjlB.jpeg",
  "vocabulary": "lvis",
  "customVocabulary": "None"
}

Output

When invoking this action, you can expect the following output:

  • image: A URL that points to a visual representation of the detected classes.
  • jsona: A URL linking to a JSON file containing detailed information about the detected classes.

Example Output:

{
  "image": "https://assets.cognitiveactions.com/invocations/7090de2b-af21-4e9e-bb24-bfe1be8e84fd/5841f7c9-24da-40c5-b7aa-61a50d65f1db.png",
  "jsona": "https://assets.cognitiveactions.com/invocations/7090de2b-af21-4e9e-bb24-bfe1be8e84fd/e41f374d-8935-467c-9d04-2e6ae988976c.json"
}

Conceptual Usage Example (Python)

Here’s a conceptual Python code snippet illustrating how to call this action:

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 = "0ca5bc57-c5c8-4995-916d-c2c29bf9f446"  # Action ID for Detect Classes Using Detic

# Construct the input payload based on the action's requirements
payload = {
    "image": "https://replicate.delivery/pbxt/K6zgH3un6xi4l07DFSUOgHKzOIbWktlvmc348VZbRDqSbqeV/k67kjlB.jpeg",
    "vocabulary": "lvis",
    "customVocabulary": "None"
}

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 snippet, you need to replace the placeholder for the API key and adjust the endpoint URL as necessary. The action ID and input payload are structured to align with the requirements of the Detect Classes Using Detic action.

Conclusion

The CudaNexus Detic Cognitive Actions provide a robust solution for developers looking to integrate sophisticated image class detection into their applications. By leveraging the power of pre-built actions, you can save time and enhance functionality without diving deep into complex machine learning models. Consider exploring additional use cases, such as real-time image analysis or automated content moderation, to further leverage these capabilities in your projects.

← Back to All Articles