This repository contains scripts for training and performing inference using an animal classification model. The model is trained on an animal dataset and can classify images and videos.

📂 Animal-Classifier
│── 📂 checkpoints/        # Directory for saving model checkpoints
│── 📂 logs/               # TensorBoard logs
│── 📂 outputs/            # Directory for saving processed videos
│── 📂 dataset/            # Dataset directory (if applicable)
│── 📄 animals_train.py    # Script for training the model
│── 📄 animals_inference_image.py  # Script for image classification inference
│── 📄 animals_inference_video.py  # Script for video classification inference
│── 📄 animals_dataset.py  # Dataset class
│── 📄 animals_model.py    # Model definition
│── 📄 requirements.txt    # Python dependencies
│── 📄 README.md           # Project documentation

• Python 3.8+
• PyTorch
• OpenCV
• NumPy
• Matplotlib
• scikit-learn
• tqdm
• TensorBoard

Run the following command to install required libraries:

pip install -r requirements.txt

To train the model, use:

python animal_train.py --data-path ./dataset --epochs 50 --batch-size 64 --learning-rate 0.001

Arguments:

• --data-path: Path to the dataset
• --epochs: Number of training epochs
• --batch-size: Batch size for training
• --learning-rate: Learning rate
• --resume: Resume training from the last checkpoint (optional)

tensorboard --logdir ./logs

To classify a single image:

python animals_inference_image.py --image-path ./test_image.jpg --checkpoint-dir ./checkpoints

Arguments:

• --image-path: Path to the input image
• --checkpoint-dir: Directory containing model checkpoints
• --checkpoint-name: Name of the checkpoint file (default: best_model.pt)

To classify animals in a video:

python animals_inference_video.py --video-path ./test_video.mp4 --checkpoint-dir ./checkpoints --show-video

Arguments:

• --video-path: Path to the input video
• --frame-size: Size to resize frames (default: 224)
• --checkpoint-dir: Directory containing model checkpoints
• --checkpoint-name: Name of the checkpoint file (default: best_model.pt)
• --show-video: Display the video with predictions in real-time

• Training logs are stored in logs/ (can be viewed using TensorBoard).
• Checkpoints are saved in checkpoints/.
• Processed videos are saved in outputs/.