Vector Similarity Search with Partially Homomorphic Encryption in Python

In this video, I demonstrate how to perform vector similarity search using partially homomorphic encryption in Python. I use the LightPHE library to build an additively homomorphic encryption algorithm and implement the Paillier encryption scheme. Alternatively, other additively homomorphic encryption algorithms can be adopted such as Exponential ElGamal, Elliptic Curve ElGamal (Weierstrass, Koblitz and Edwards forms), Paillier, Damgard-Jurik, Okamoto–Uchiyama, Benaloh, Naccache–Stern. I conduct an experiment in facial recognition by representing facial images as vector embeddings using DeepFace and its default model VGG-Face. This approach enables secure similarity search while preserving privacy by performing computations on encrypted data. Tutorial: https://sefiks.com/2025/03/04/vector-similarity-search-with-partially-homomorphic-encryption-in-python/ Please Subscribe! That's what keeps me going ► https://bit.ly/40NfIS7 Want more? Connect with me here: Blog: https://sefiks.com/ Twitter: https://twitter.com/serengil Instagram: https://www.instagram.com/serengil Facebook: https://www.facebook.com/sefikscom Linkedin: https://www.linkedin.com/in/serengil/ If you do like my videos, you can support my effort with your financial contributions on - Patreon: https://www.patreon.com/serengil?source=youtube - GitHub Sponsors: https://github.com/sponsors/serengil - Buy Me a Coffee: https://buymeacoffee.com/serengil