/reverse-image-search

An image search implementation in python using tensorflow keras, scikit-learn, scipy and matplotlib.

Primary LanguagePython

Hy Image Search

An image search implementation in python using tensorflow keras, scikit-learn, scipy and matplotlib.

The image embeddings are generated from Xception imagenet (can be changed/tuned from features.py). Currently the embeddings are stored using pickle, but a database may be used instead. Image embeddings are compared cosine similarity and hamming.

Running search on image(s) from console args

  1. Run search.py with image path(s) as arguments and the script will display the top matches from underlying image library

Sample

Running search on a custom set of images

  1. Transfer the images to queries folder data/queries/ or alternatively change query_images_folder_path from paths.py to your location of your image set.
  2. Run search.py and the script will display the top matches from underlying image library

Sample

Generating embeddings for a new image set

  1. Transfer the images to images folder data/images/ or alternatively change images_folder_path from paths.py to your location of your image set.
  2. Run features.py and you should have the embeddings generated

Viewing the similarity map in underlying image library (PCA and t-SNE)

  1. Run dataset.py and you should get a visualization similar to this

The sample images used here are from idenprof dataset