Skip to content

This repository is part of a blog post that guides users through creating a visual search application using Amazon SageMaker and Amazon Elasticsearch service

License

Notifications You must be signed in to change notification settings

alexvt-amz/amazon-sagemaker-visual-search

 
 

Repository files navigation

Visual image search

This repository is part of AWS blog to create a visual search application using Amazon SageMaker and Amazon Elasticsearch

How does it work?

We will use Fashion Images from feidegger, a zalandoresearch dataset as a reference image to generate a 2048 feature vector using a convolutional neural networks and gets stored into Amazon Elasticsearch KNN index

diagram

When we present a new query image, it's computing the related feature vector from Amazon SageMaker hosted model and query Amazon Elasticsearch KNN index to find similar images

diagram

Use Keras

Refer to src and visual-image-search.ipynb

Use GluonCV

Refer to endpoint and visual-image-search-gluoncv.ipynb, train script is transfer_learning.ipynb

License

This library is licensed under the MIT-0 License. See the LICENSE file.

About

This repository is part of a blog post that guides users through creating a visual search application using Amazon SageMaker and Amazon Elasticsearch service

Resources

License

Code of conduct

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 61.4%
  • Python 21.1%
  • JavaScript 12.5%
  • HTML 1.5%
  • Dockerfile 1.4%
  • Shell 1.4%
  • CSS 0.7%