CNN-based Image Retrieval System

Abstract — The goal of this paper is to develop an image search and similarity founding system using convolutional neural network model. The implementation of this model will be done using the TensorFlow and NumPy libraries. When using a regular search engine, text search can often provide a wide range of products, but they may not be similar to what we are looking for. This can be frustrating and time-consuming. A better solution is to use content-based image retrieval technology, which allows us to retrieve similar products based on an image query. This system works by extracting features from all images in the database, including the query image, using a feature extraction algorithm. Similarities are then calculated between the query image and all images in the database, and the system retrieves all images that have a high degree of similarity with the query image. This technology provides a more efficient and accurate way of finding products that match our preferences.