Logo Lanfrica

Waqar-Abid/WildNet-Animal-Image-Classifier

Domaine:

environment and energy

Type de record:

projectmodel
Créateur:
Waq
Hôte:
WildNet Animal Image Classifier 🦁📸 A deep learning project using Convolutional Neural Networks (CNNs) for image classification of African wildlife (e.g., elephant, giraffe, lion, zebra). Built with PyTorch and trained on a custom dataset as part of a computer vision challenge. # 🦁 WildNet Animal Image Classifier This project implements a Convolutional Neural Network (CNN) using PyTorch to classify African animals (e.g., Lion, Elephant, Zebra, Giraffe) based on images. The model is trained and evaluated on a custom dataset. --- ## 🚀 Features - Built using PyTorch with a simple CNN architecture - Supports training, validation, and testing splits - Tracks accuracy and loss over epochs - Saves trained model to `.pt` file - Inference visualization using Matplotlib --- ## 🧠 Model Summary - **Convolutional Layers:** 3 - **Activation:** ReLU - **Pooling:** MaxPool - **Fully Connected Layers:** 2 - **Optimizer:** Adam - **Loss Function:** CrossEntropyLoss - **Validation Accuracy:** 100% --- ## 📦 Requirements Install required libraries: ```bash pip install -r requirements.txt ``` --- ## 🏃‍♂️ How to Run ### ▶️ Train the Model ```bash # Make sure you're in the project folder jupyter notebook ``` - Open: `WildNet Animal Image Classifier.ipynb` - Run all cells sequentially to train the model --- ### 🧪 Inference on Test Set After training completes, the notebook visualizes predictions on random test images from the test set. --- ## 💾 Model Saving The trained model is automatically saved at: ```bash models/safari_classifier.pt ``` You can later load this model for inference. --- ## 🙋‍♂️ Author **Waqar Abid** WildNet Animal Image Classifier --- ## 📬 Contact - LinkedIn: waqarabid - GitHub: @waqar-ai