/Wild-Animal-Classifier

A Deep Learning Project to classify images of wild animals into 10 classes where Fine Tuning Technique has been adopted for better model performance.

Primary LanguageJupyter Notebook

Wild-Animal-Detection

Computer vision-based system for the detection of wild animals

Installation:-

  1. Python 3.6 or above
  2. Tensorflow 2.4 or above
  3. Numpy 1.2 or above
  4. Matplotlib 3.1 or above
  5. Pandas 1.4 or above

Configuration Instructions:-

  1. Download the dataset
  2. Store the different classes of Animals into different folders here 10
  3. Do preprocessing using KMeans Segmentation
  4. Do artifical sythesis of Images using Data Augmentation to solve class imbalance
  5. Run the Wild_Animal_Detection.ipynb file as a Jupyter Notebook or Colab Notebook

Hardware Requirements:-

  1. RAM: 16GB or above
  2. Graphics: RTX 3050 or above 4GB+ VRAM

Software Requirements:-

  1. OS: Windows or Mac OS or Linux

Path:- train_data_path is the path to training data and valid_data_path is the path to testing data