object-detection

Unlocking Object Detection with YOLOX: A Developer's Guide

26 Apr 2025
Unlocking Object Detection with YOLOX: A Developer's Guide

In the rapidly evolving domain of computer vision, the ability to detect and identify objects within images is pivotal for a range of applications, from autonomous vehicles to security systems. The YOLOX service provides a robust solution for developers looking to integrate object detection capabilities into their applications. By leveraging the YOLOX model, you can achieve high-speed, accurate object detection with customizable parameters to suit your needs.

This article will delve into the capabilities of YOLOX's object detection action, detailing its features, input requirements, expected outputs, and various use cases. Whether you're building an image processing application, enhancing security surveillance, or developing an augmented reality experience, YOLOX offers the tools you need to succeed.

Prerequisites

Before diving into the YOLOX object detection action, ensure you have a Cognitive Actions API key and a basic understanding of making API calls. This will allow you to seamlessly integrate the service into your projects.

Detect Objects Using YOLOX

The "Detect Objects Using YOLOX" action empowers you to utilize the advanced YOLOX model for detecting objects in images. This action is essential for developers aiming to implement object recognition features in their applications.

Purpose

The primary purpose of this action is to detect and classify multiple objects within an input image by using various YOLOX model versions. You can adjust parameters such as confidence thresholds and non-maximum suppression (NMS) thresholds to optimize detection results based on your specific requirements.

Input Requirements

To use this action, you must provide the following inputs:

  • inputImage: A URI path to the input image (required).
  • modelName: Choose from available model sizes ("yolox-s", "yolox-m", "yolox-l", "yolox-x"). The default is "yolox-s".
  • confidenceThreshold: Set the confidence level for detections. Only detections with a confidence higher than this threshold will be retained. The default value is 0.3.
  • nmsThreshold: This threshold controls the overlap percentage for detections considered redundant. The default is 0.3.
  • targetSize: Resize the input image to this size (default is 640 pixels).
  • returnJson: Optionally, specify whether to return results in JSON format (default is false).

Expected Output

When you call this action, the expected output will include:

  • img: A URI to the image with detected objects highlighted.
  • json_str: If returnJson is set to true, this will provide a JSON representation of the detected objects (otherwise, it will be null).

Use Cases for this Specific Action

  • Security and Surveillance: Enhance security systems by integrating real-time object detection to identify potential threats or monitor specific areas.
  • Retail Analytics: Implement object detection in retail environments to analyze customer behavior and improve inventory management.
  • Autonomous Vehicles: Use YOLOX for detecting pedestrians, vehicles, and obstacles in real-time, aiding navigation and safety features.
  • Augmented Reality: Create immersive experiences by detecting and interacting with real-world objects in augmented reality applications.
import requests
import json

# Replace with your actual Cognitive Actions API key and endpoint
# Ensure your environment securely handles the API key
COGNITIVE_ACTIONS_API_KEY = "YOUR_COGNITIVE_ACTIONS_API_KEY"
# This endpoint URL is hypothetical and should be documented for users
COGNITIVE_ACTIONS_EXECUTE_URL = "https://api.cognitiveactions.com/actions/execute"

action_id = "dab86e16-b55e-4857-8e4b-8934545141b4" # Action ID for: Detect Objects Using YOLOX

# Construct the exact input payload based on the action's requirements
# This example uses the predefined example_input for this action:
payload = {
  "modelName": "yolox-s",
  "inputImage": "https://replicate.delivery/pbxt/IklFY8BBmH3uU1kSooUnsglSh0UvNf4EsMTfrgK6K8R30gof/office.jpg",
  "returnJson": false,
  "targetSize": 640,
  "nmsThreshold": 0.3,
  "confidenceThreshold": 0.3
}

headers = {
    "Authorization": f"Bearer {COGNITIVE_ACTIONS_API_KEY}",
    "Content-Type": "application/json",
    # Add any other required headers for the Cognitive Actions API
}

# Prepare the request body for the hypothetical execution endpoint
request_body = {
    "action_id": action_id,
    "inputs": payload
}

print(f"--- Calling Cognitive Action: {action.name or action_id} ---")
print(f"Endpoint: {COGNITIVE_ACTIONS_EXECUTE_URL}")
print(f"Action ID: {action_id}")
print("Payload being sent:")
print(json.dumps(request_body, indent=2))
print("------------------------------------------------")

try:
    response = requests.post(
        COGNITIVE_ACTIONS_EXECUTE_URL,
        headers=headers,
        json=request_body
    )
    response.raise_for_status() # Raise an exception for bad status codes (4xx or 5xx)

    result = response.json()
    print("Action executed successfully. Result:")
    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 (non-JSON): {e.response.text}")
    print("------------------------------------------------")

Conclusion

Integrating the YOLOX object detection action into your applications can significantly enhance their capabilities, allowing for real-time analysis and interaction with the visual world. With customizable parameters and multiple model options, YOLOX offers flexibility to meet various use cases, from security to retail analytics.

Now that you understand the benefits and applications of YOLOX, consider exploring additional features or integrating it into your next project for a powerful object detection solution.

← Back to All Articles