Using Keras and VGG16(pre-trained model already trained on ImageNet dataset) used for transfer learning.
- Install Anaconda Python Distribution.
- Install various python libraries
# pip install -r requirements.txt
- Run command prompt,
> jupyter notebok
Clone the repo
# git clone https://github.com/A4ANK/face_recognition_Transfer_Learning.git In repo directory, create two sub directories
# mkdir train
# mkdir test Similarly, also create subdirectories inside test and train directories for subcategories that you want to predict, respectively. Now, collect dataset using face_extractor.ipynb notebook
Load the VGG16 pre-trained model inside Transfer Learning.ipynb Use transfer learning to train this pre-trained model on the new smaller dataset created using face_extractor.ipynb notebook.
Run Real Time Face recognizer.ipynb notebook to perform.