CBIR model which has been tested with multiple feature descriptors such as SIFT, SURF, ORB, BRIEF and combinations of them. Multiple linguistic processing techniques such as Bag of Words and Topic modelling have been used for optimizing the image retrieval and making them meaningful based on human semantics. Dataset link is https://www.kaggle.com/jessicali9530/caltech256 The final.zip provided has the flask application for image retrieval the pickle files are also provided which can be generated using the jupyter notebook provided the subset of the dataset is also provided.