Developing a database to store 100-dimensional image embeddings along with their paths

Suppose we want to design a database to store images represented in a vector format (each image is stored as a 100-dimensional vector). The goal here is to store image embeddings along with their paths so that we can use the embedding to find image neighbors, which helps in image retrieval. Once we find the neighbors to images, I would like to return back and display images, and that is why I need to store image paths as well in the database. The KD-tree data structure is good to start for the nearest neighbors search, but it suffers from dimensionality issues given we have a vector of dimension equal to 100. **First question**: Can you please let me know what you think about this design strategy and what would you recommend as an alternative to KD-tree to deal with high dimensional input? **Second question**: Is it easy to find sophisticated nearest neighbor search in Oracle database for example?