CoolFace
Apppublic

rnunez245/helmet-detection-backend

sourceHugging Facemitupdated 8mo agoView on Hugging Face
0likes
App README

Helmet Detection Computer Vision - Backend API

Flask REST API for safety helmet detection using VGG-16 transfer learning.

API Endpoints

GET /

Health check and API documentation

POST /v1/predict

Detect helmet presence in uploaded image

Input: multipart/form-data with field name "image" Accepts: JPG, JPEG, PNG (max 10MB) Returns: JSON with class, confidence, and label

Example:

bash
curl -X POST https://rnunez245-helmet-detection-backend.hf.space/v1/predict \
  -F "image=@test_image.jpg"

Response:

json
{
  "class": 1,
  "confidence": 0.9987,
  "label": "Helmet Detected",
  "model_version": "1.0"
}

Model Details

  • —Architecture: VGG-16 (ImageNet) + Feed-Forward Neural Network
  • —Accuracy: 100% on test set
  • —Input: 200×200 RGB images
  • —Output: Binary classification (helmet/no helmet)

Links

Note

⚠️ Model file will be added soon. API is live but predictions require the trained model file.