Image Retrieval System

Implemented a sophisticated Content-based Image Retrieval System that excels in finding similar images across various perspectives, such as Street vs. Satellite or Ground vs. Aerial views.

For the retrieval system, I created a pipeline that efficiently processes a query image and returns well-organized and sorted results that match the query image, akin to the functionality provided by Google Image Search.

To accomplish image search, I leveraged a fine-tuned VGG16 model for classification, employing an adversarial approach to learn and extract essential image features effectively. Google Colaboratory and Tensorflow have been utilized in training and testing the model.