Engineering Considerations for Running Machine Learning Models at the Edge: Application in Body Scanning for eCommerce

This talk will review the engineering considerations to support the operation of body scanning machine learning models to evaluate the usability of images for generating a body double in the Amazon shopping application. In order to create next generation shopping experiences which reflect a customer’s body proportions based on their photos, the engineering team needed a low latency way to validate if the data would be able to produce a highly accurate result. The body estimation models were used to generate a 3D simulation of a custom shirt draped on the person using the iOS or Android shopping app. Engineering requirements for operating this model at the edge included latency of application download, model run time, model size, and the ability to instrument the experience for defection detection while keeping the customer data secure.


Speaker

Jenn Lin

Principal Engineer and Sr. Software Development Manager @Amazon

Jenn Lin is currently a Principal Engineer and Sr. Software Development Manager leading a team for Amazon's Selling Partner Services team that focuses on automated profitability for the Amazon business. Prior to this, Jen was a Principal Engineer for Amazon Fashion and building the next generation of clothes shopping experiences.

Read more

Speaker

Herak Sen

Principal Software Engineer @Amazon

Herak Sen is currently a Principal Engineer for Amazon Fashion where he focusses on helping customer find the perfect fit. He is responsible for building next generation of customer shopping experiences and garment manufacturing technologies.

Read more

Date

Tuesday Oct 25 / 11:50AM PDT ( 50 minutes )

Location

Marina

Topics

Machine Learning Database Engineering Development

Share

From the same track

Session Kubernetes

Kubernetes and LaunchDarkly; The Junction of Deploy and Software

Tuesday Oct 25 / 01:40PM PDT

Teams are leveraging Kubernetes (and other container based technologies) to improve the speed at which they ship and deploy applications, however, Kubernetes still largely focuses on solving these problems through infrastructure concepts.

Speaker image - Peter McCarron

Peter McCarron

Senior Technical Marketing Engineer @LaunchDarkly

Session Data Analytics

Understanding Analytics and Data-driven Decision Making

Tuesday Oct 25 / 02:55PM PDT

Understanding Analytics and Data-driven Decision Making

Speaker image - Daniel Ceasar Paul  Jalathyan

Daniel Ceasar Paul Jalathyan

Application Performance Management @Zoho

Session Java

Performance Testing Java Applications

Tuesday Oct 25 / 10:35AM PDT

Every so often, you’ll read a performance benchmark (of a Java or other application), with bold claims for how well X performs compared to Y.

Speaker image - Pratik  Patel

Pratik Patel

Java Champion & developer advocate @Azul Systems

Session Programming

Discover Inspirational Insights in Motivational Sports Speeches Using Speech-to-Text

Tuesday Oct 25 / 05:25PM PDT

Inspirational sports speeches have motivated and reinvigorated folks for years. Whether you’re a developer or an athlete, they’ve withstood the journey because even the smartest, the bravest, and the most resilient need some encouragement on occasion.

Speaker image - Tonya  Sims

Tonya Sims

Python Developer Advocate @Deepgram

Session Programming

Are Programming languages... *Actually* Languages?

Tuesday Oct 25 / 04:10PM PDT

Spoiler: proooooooobably not… But Second Language Acquisition and Tech Skill Building share a *lot* of similarities. Despite their best efforts, people regularly stumble in their language-learning AND Dev/DevOps journeys. Why is this, and what can we do better?

Speaker image - Dylan Lacey

Dylan Lacey

Manager of Developer Relations @Sauce Labs