This project houses two implementations of recognizing human faces, particularly those present in the publicly available 'LFW' dataset (more than 13,000 images; faces recognized with facenet-pytorch
) and 'Indian Actor Faces' dataset (more than 5,000 images; faces recognized with face_recognition
).
- Developed using VS Code, Jupyter Notebook and Google Colab
- Python and it's libraries - Flask, Pytorch, face_recognition, pickle, faiss, Pillow, opencv and more
match_face
andrecognise_actor
APIs tested with Postman
Potential improvements that I want to implement:
- Implementing a Vector DB, like
Milvus
for practical applications like managing attendance - Using libraries like
deepface
- Handling multiple persons in an image
- Adding more information for better face matching!
Thanks to Akshay sir, Abhishek sir, Balesh sir, and Mayank for the constant guidance and support.
Thanks to Cubastion Consulting Pvt. Ltd. for a productive and supportive environment that fosters learning.