Documentation

Everything you need to know about using ImageAI

Getting Started

How to Use the App

  1. Navigate to the Classify page from the navigation menu
  2. Select one of the available AI models from the dropdown
  3. Upload an image by clicking the upload area or dragging and dropping
  4. Click the Predict button to start classification
  5. View your prediction results along with confidence scores

Tip: For best results, use clear, well-lit images with the subject in focus. Supported formats include JPG, PNG, and GIF.

Available Models

Choose from four state-of-the-art deep learning models

ResNet-34

Fast

A residual neural network with 34 layers. Offers a great balance between speed and accuracy.

  • Accuracy: 98.0%
  • Best for: Quick predictions

DenseNet-121

Medium

Dense convolutional network with 121 layers. Features dense connections between layers for improved gradient flow.

  • Accuracy: 98.0%
  • Best for: General purpose classification

EfficientNet-B7

Slow

The most accurate model using compound scaling. Best when accuracy is the top priority.

  • Accuracy: 99.0%
  • Best for: Maximum accuracy

InceptionV3

Medium

Uses inception modules with multiple filter sizes. Excellent for diverse image types.

  • Accuracy: 96.0%
  • Best for: Varied image content

API Integration

Technical details for developers

Prediction Endpoint

POST

http://localhost:5000/predict

Request Format

{
  "model": "resnet34",
  "image": "base64_encoded_image_data"
}

Response Format

{
  "prediction": "golden_retriever",
  "confidence": 95.8,
  "model": "resnet34"
}

Supported Model Values

  • resnet34
  • densenet121
  • efficientnet-b7
  • inceptionv3

Note: The Flask backend must be running on http://localhost:5000 for predictions to work. Ensure your backend server is properly configured and running before making predictions.