/custom_face_recognition

Face Recognition using SVM and RF is a simple image classification model for celebrity images. The model has been trained on 3 celebrities with 100 images each, achieving 71% accuracy with SVM and 68% accuracy with RandomForest algorithms

Primary LanguageJupyter Notebook

Face Recognition using SVM and RF

Hi

This is an simple image classification model for celebrity images

The model I created has been trained 3 celebrity images with 100 images each I have used SVM and RandomForest algorith to perform the training with 71% and 68% accuracy respectively

to run it on your personal system

  • Step 1: download all the files in this repo
  • Step 2: go to cmd and type pip install -r requirements.txt
  • Step 3: type juypter-notebook in the same directory as the files(assuming that jupyter notebook is already present)
  • Step 4: open recognizer.ipynb file in jupyter notebook and run all the cells
  • Step 5: You can input your own dataset by typing the name of celebrities in the cell that takes input and downloads images from binge image downloader or you can directly paste ur labeled directory that has images of the celebrity in images folder
  • Step 6: run all the cells
Accuracy metrics : 1) accuracy score 2)confusion matrix