SnapMode: First Iranin Large-Scale Image Retrieval System

It presents a distributed stream processing platform for the real-time analysis of fashion data based on open-source big data analysis tools including Apache Storm, Apache Kafka, Apache Solr.
It proposes a disentangled feature extraction method through a deep convolutional generative adversarial networks (DCGANs) for content-based image indexing and retrieval.
It provides a solution to power embedding similarity search through the Milvus engine

Please visit the system using the following URL:

Unfortunately, due to the commercial nature of the product, I can not share its code. Please Find the paper at this URL.

system architecture

**Overall view of system architecture **

Overview of two main subsystems: (i) Web Crawler and (ii) Image Processing

Overview of Crawler Subsystem

Overview of Image Processing Subsystem

The sample screenshot of the SnapMode Web-based Application.