Implement of numerical feature vectors for an image dataset and retrieve based on it. To do this, four distance functions(Eucleadian Distance, Cosine Distance, Minkowski Distance and Jaccard Distance) are considered and then using each of these functions, the image numerical vector is created and the retrieval operation is performed. For this project, a collection of images of Oxford buildings has been used. Eleven standard queries have been defined for this dataset, and the relationship between documents and queries has been identified. You can see jupyter notebook of this project in repository or you can access it directly from this link. Also you can use this google drive folder to access required directories in notebook.
mobinamosannafat / distance-based-search-engine Goto Github PK
View Code? Open in Web Editor NEWImplement of numerical feature vectors for an image dataset and retrieve based on it