Real-Time Fire & Smoke Detection API

Empower your applications with our state-of-the-art YOLOv5 computer vision model. Detect flames and smoke instantly using our fast, reliable REST APIs or test it live in your browser.

Try Live Demo Read API Docs
Fire Detection Example
Hover to view API metadata
Model Metadata
{
  "summary": {
    "total_detections": 1,
    "flame_count": 1,
    "smoke_count": 0,
    "highest_confidence": 0.8777340650558472
  },
  "metadata": {
    "width": 640,
    "height": 640,
    "format": "JPEG"
  },
  "predictions": [
    {
      "xmin": 322.4617004394531,
      "ymin": 23.019607543945312,
      "xmax": 374.1841735839844,
      "ymax": 164.92213439941406,
      "confidence": 0.8777340650558472,
      "class": 1,
      "name": "flame"
    }
  ]
}

Lightning Fast

Optimized YOLOv5 architecture ensures minimal latency for real-time video stream processing and rapid image analysis.

Developer Ready

Simple REST APIs returning rich JSON payloads or annotated images. Fully documented with interactive Swagger and ReDoc support.

Highly Accurate

Trained on thousands of diverse environments to minimize false positives while ensuring reliable fire and smoke detection.

Model Architecture

Our AI team at Detectium has developed a cutting-edge computer vision API using YOLOv5 to detect flames and smoke in real-time. This model utilizes our innovative method of training machine learning models on diverse datasets, yielding robust results across different environmental conditions.

At Detectium, we believe in continuous development. We are actively incorporating feedback from developers and users to iteratively refine and expand our model's capabilities. Thank you for your continued support!