Skip to content

Latest commit

 

History

History
43 lines (26 loc) · 1.76 KB

README.md

File metadata and controls

43 lines (26 loc) · 1.76 KB
Raw

Celebrity-Face-Recognition

The Deep Learning Model is trained using Tensorflow Framework and it gives validation accuracy value of 85%

Dataset Images

Description:

Download the dataset from the kaggle link given below the description and split it into the training set and and validation set based on the given ratio.Transfer Learning has been implemented with the MobileNetV2 model and imagenet weights. The pre-trained model consists of 155 layers and the whole layers of the pretrained model is retrained to get the better accuracy.Image Augmentation has been implemented to avoid the overfitting of the model.The model has been trained for the 50 epochs and the model has scored Validation Accuracy of 85.

Dataset Content:

This images has been collected from Pinterest and cropped. There are 105 celebrities and 17534 faces

Prerequisites:

The below libraries are needed to execute the Python code

  • Python 3.8
  • Tensorflow 2.1.0
  • Numpy
  • Matplotlib

Dataset Link: https://www.kaggle.com/hereisburak/pins-face-recognition

Kaggle Link: https://www.kaggle.com/srikeshram/celebrity-face-recognition

Pre-Trained Weights:

The pre-trained model weights for the classification of 105 celebrities has been uploaded as a .zip file which can be extracted to obtain the .h5 file. This file can be loaded with the Tensorflow in case of without training the model again to obtain the new weights.

Repository Contains:

  • Python Code (.ipynb)
  • Trained Weights(.h5)

License:

This project is licensed under the Apache License 2.0- see the LICENSE.md file for details

Acknowledgements:

  • hereisburak for Dataset
  • Inspiration
  • etc...