Created the database and enhanced the camera functionality.

[kamisama-coder]

If a user enters two names, the system will first check if they are already present in the database. If not, it will store the corresponding embeddings along with the names. The cosine similarity function will then compare these embeddings with those of the images stored in the database, considering a similarity range between 0.5 and 1.

Additionally, I plan to optimize the clustering process based on stored embeddings and names. This will enhance efficiency and allow the system to retrieve images containing not just the specified individuals but also others who appear alongside them.

[kamisama-coder]

@Pranav0-0Aggarwal pls review the PR