Intelligent Image Classification with KNN

This n8n workflow provides a robust solution for intelligent image classification. It begins by generating multimodal embeddings for an input image using Voyage AI's advanced models. These embeddings are then used to query a Qdrant vector database, searching for the most similar image embeddings. A K-Nearest Neighbors (KNN) approach is employed to retrieve the top-ranked neighbors. A 'Majority Vote' mechanism then determines the most probable class based on the labels of these neighbors. The workflow includes logic to handle ties and iteratively increase the search radius if necessary, ensuring more accurate classification results.