This repository provides resources for implementing a visual search engine. Visual search is the central component of an interface where instead of asking for something by voice or text, you show what you are looking for.
When shown a real world, physical item, an AWS DeepLens device generates a feature vector representing that item. The feature vector generated by the AWS DeepLens device is sent to the AWS cloud using the AWS IoT service. An AWS IoT Rule is used to direct the incoming feature vector from the device to a cloud-based AWS Lambda function, which then uses that feature vector to search for visually similar items in an index of reference item feature vectors. This index is created using SageMaker's k-Nearest Neighbors (k-NN) algorithm. The search Lambda function returns the top visually similar reference item matches, which are then consumed by a web app via a separate API Lambda function fronted by Amazon API Gateway.
