Sign In

Edward Miller

Product Manager, Computer Vision Research at Meta

Edward Miller is a tech professional serving as the site lead for the Reality Labs research team in London, focusing on scene understanding technologies and overseeing open-source initiatives like Project Aria. He has a history of successfully transferring research outcomes to tangible products.

Edward Miller's background includes co-founding an AI company specializing in precise location understanding via computer vision, which was acquired by Facebook. He led a team of 30 individuals in this venture.

Educationally, Edward studied Exponential Medicine at Singularity University and holds a Bachelor of Arts in Philosophy & Psychology from The University of Sheffield. His diverse academic background complements his tech expertise well.

Throughout his career, Edward Miller has held key roles in various organizations including Meta's Reality Labs Research, Scape Technologies (acquired by Facebook), Medical Realities, Immersiv.ly, and as a VR & Interactive Imagery Specialist at Independent. His professional journey reflects a deep involvement in cutting-edge technology and research-driven ventures.

Highlights

Aug 31 · linkedin

📣📣📣 Introducing EgoBlur, a new open source model from the team behind Project Aria at Reality Labs Research.

EgoBlur helps to preserve bystander privacy by detecting PII in images and videos so that the contents can be blurred.

Used internally by Project Aria, Meta’s device for conducting egocentric research, EgoBlur is used to detect and then remove faces and license plates from RGB or grayscale images that have been captured in unconsented environments.

Working with external academic partners, we saw researchers struggle with the existing tools for removing personally-identifiable information from research datasets - particularly when working with data from an egocentric perspective (i.e. first-person perspective)

As such, our team has released the EgoBlur models under the Apache 2.0 license. This means that the models and tools can be used to help enable responsible innovation not only in academia, but also in industry.

We’re looking forward to seeing how EgoBlur will empower teams from academia and industry alike to accelerate the state of research for AR and AI applications, responsibly.

To access the models, read the paper, and learn more about the EgoBlur models, visit https://lnkd.in/eCDpNe_2.

Edward Miller
Add to my network

Location

London Area, United Kingdom