ishan-dhakal19/rabbit_detector

🐰 Rabbit Detector AI

AI-powered rabbit detection using deep learning and data augmentation.

Features

• —🤖 Deep Learning: Trained CNN model for rabbit detection
• —📊 High Accuracy: 85-90% detection accuracy
• —⚡ Fast Inference: Results in seconds
• —🎨 Data Augmentation: Trained from single image with augmentation

How It Works

This model was trained using data augmentation technique:

1. 1.Started with 1 rabbit image
2. 2.Created 200 augmented variations (rotations, flips, zooms, brightness)
3. 3.Generated 200 negative samples
4. 4.Trained a CNN with 3 convolutional layers
5. 5.Achieved 85-90% validation accuracy

Technology Stack

• —Framework: TensorFlow/Keras
• —Frontend: Streamlit
• —Model: Custom CNN (500K parameters)
• —Input Size: 128x128 RGB images

Usage

Simply upload an image and the AI will:

• —✅ Detect if a rabbit is present
• —📈 Show confidence score
• —📊 Display technical details

Model Architecture

Input (128x128x3)
    ↓
Conv2D(16) + MaxPool + Dropout
    ↓
Conv2D(32) + MaxPool + Dropout
    ↓
Conv2D(64) + MaxPool + Dropout
    ↓
Dense(128) + Dropout
    ↓
Output (Sigmoid)

Try It!

Upload any image to test the detector!

Credits

Built with ❤️ using TensorFlow and Streamlit