Technology

Technology

AI Powered Authentication

Entrupy’s intelligent, network-driven solutions for product authentication are powered by artificial intelligence, with proprietary algorithms trained on a rich and diverse data set of millions of images from known-authentic and known-counterfeit products. These algorithms run security, in the cloud, ensuring complete objectivity and eliminating risks of human error and bias.
When an item is processed using our app, a set of images are collected and analyzed by these algorithms, which run securely in the Cloud. With speed and pinpoint accuracy, the algorithms either verify the item’s authenticity or return an “unverified” result within minutes and often instantly. Each new scan adds to the database in real time, further training the algorithms and making our solution smarter with every verification performed.

AI POWERED AUTHENTICATION

Entrupy’s secure, accurate and easy to use solutions are built for fast implementation and unlimited scale, require only a few components:
  • Portable Device that magnifies the images 260x with uniform lighting (Luxury Authentication only)
  • Mobile app that guides the user through the process of image collection
  • Machine learning algorithms, operating securely in the cloud.
The solution uses 256 bit encryption, securing Edge to Cloud control and data traffic

Data and Algorithms: Foundations of Our Authentication Solution

Entrupy trains its algorithms to differentiate between fake and authentic items using an unique set of data with millions points based on known-real and known-counterfeit items. Because the solution is machine learning-driven, every item that undergoes verification helps the algorithms learn and improve, creating a smarter solution that adapts to the changing world.

Data Sources

Partnering with some of the most prominent and trustworthy businesses around the world, our data teams have collected millions of images from a variety of categories of physical goods (bags & accessories, shoes, watches), building a data set that is rich, diverse and completely unmatched.

Data Quality

To ensure our data meets the highest quality standards, we run all submitted data through three layers of filtering, two manual and one algorithmic. This enables us to correct any possible errors while guaranteeing the algorithms are fed only extremely high-quality information during the machine-training process.

Data Processes

Entrupy’s data includes styles created between 80 years ago and today. Our data teams track new releases from brands, capturing and incorporating their data into our solutions as soon as possible, giving customers the ability to perform verification services across the widest possible time range for supported brands.

To learn more about the data science that makes Entrupy’s authentication system possible, please check out the following white paper written by Entrupy’s co-founders and published by The Association for Computing Machinery (ACM).

Check out this paper published by the founders of Entrupy to learn more about the technology powering our authentication system.

The Fake vs Real Goods Problem: Microscopy and Machine Learning to the Rescue

Ashlesh Sharma (Entrupy Inc); Vidyuth Srinivasan (Entrupy Inc); Vishal Kanchan (Entrupy Inc); Lakshminarayanan Subramanian (Entrupy Inc)

Entrupy published on KDD.org

“We introduce a new mechanism that uses machine learning algorithms on microscopic images of physical objects to distinguish between genuine and counterfeit versions of the same product. The underlying principle of our system stems from the idea that microscopic characteristics in a genuine product or a class of products (corresponding to the same larger product line), exhibit inherent similarities that can be used to distinguish these products from their corresponding counterfeit versions. A key building block for our system is a wide-angle microscopy device compatible with a mobile device that enables a user to easily capture the microscopic image of a large area of a physical object. Based on the captured microscopic images, we show that using machine learning algorithms (ConvNets and bag of words), one can generate a highly accurate classication engine for separating the genuine versions of a product from the counterfeit ones; this property also holds for super-fake counterfeits observed in the marketplace that are not easily discernible from the human eye. We describe the design of an end-to-end physical authentication system leveraging mobile devices, portable hardware and a cloud-based object verication ecosystem. We evaluate our system using a large dataset of 3 million images across various objects and materials such as fabrics, leather, pills, electronics, toys and shoes. The classication accuracy is more than 98% and we show how our system works with a cellphone to verify the authenticity of everyday objects.
To download and read the complete paper, please click here.