Computer Vision · Graphics · Applied AI

Computer vision for the world's hardest perception problems.

Flagg Visual Computing is a collective of computer-vision, graphics and AI specialists. We embed with your team to design, prototype and ship production perception systems — from autonomous racecars to biometric identity.

Ph.D., Georgia Tech Chief Science Officer, Code19 Racing 15+ clients shipped
Matt Flagg, Ph.D. Matt Flagg, Ph.D.

Principal & Founder

Matt Flagg received his Ph.D. in Computer Science from Georgia Tech in 2010, specializing in computer vision, graphics and human-computer interaction. His tracking research earned the Best Student Paper award at the British Machine Vision Conference and a Foley Scholars Award, and his peer-reviewed work — spanning projector-camera systems, motion-coherent tracking and video synthesis — has been cited more than 1,100 times in venues including IJCV, CVPR and ACM UIST.

Before consulting, he co-founded PlayMotion, a gesture-recognition company that deployed hundreds of projector-camera systems across hospitals, museums and family entertainment centers, including three divisions of Disney parks. He also contributed to a 3D scanning system honored with a Scientific and Engineering Award from the Academy of Motion Picture Arts and Sciences at Industrial Light & Magic. That early work grew into a portfolio of more than ten U.S. patents covering gesture control, interactive displays, virtual try-on, active lighting and identity-document security.

Over the past decade Flagg has researched and developed perception systems for more than fifteen clients, from Meta Reality Labs to venture capital investors. His engagements span the full breadth of visual computing — LiDAR-based opponent detection for autonomous racecars, real-time markerless motion capture of professional soccer matches, document authentication and biometric fraud detection, 3D facial scanning for cosmetic pre-consultation and custom eyewear, photo-to-avatar synthesis that helped launch AR Emoji on Samsung devices, and technical due diligence of computer-vision investments.

Today Flagg is Chief Science Officer at Code19 Racing — the world's first professional autonomous motorsports team — and leads Flagg Visual Computing, a collective that pairs him with a deep bench of longtime engineers and researchers. He collaborates with longtime partner Dr. Satya Mallick (CEO of OpenCV), embedding with each client's team to design, prototype and ship perception systems end to end. All client IP is owned by the customer.

15+Clients shipped
2010Ph.D. Georgia Tech
1stPro autonomous race team
3Disney park divisions

What we do

Perception systems, end to end.

Expertise and software-development firepower in computer vision, graphics and AI — available for hire on an hourly or fixed-bid basis.

Autonomous Perception

LiDAR, radar and camera fusion, 3D point-cloud perception and multi-object tracking for vehicles operating at the friction limit.

Identity & Documents

Document authentication and fraud detection for secure identity verification at scale.

3D Reconstruction & Scanning

3D face and object scanning, underwater reconstruction and photogrammetry — from cosmetic pre-consultation to custom-printed eyewear.

Markerless Motion Capture

Real-time, markerless tracking of athletes for professional sports broadcast and analytics — including live soccer matches.

Avatars & Generative Graphics

Photo-to-avatar creation and video synthesis — work that helped launch AR Emoji on Samsung devices.

Applied AI & Due Diligence

Visual search, AI-driven video compression and technical due diligence of computer-vision investments for venture and PE.

Selected work

Clients & engagements.

Most engagements involve multi-year retention of Flagg and his team. A sample of the companies, labs and investors we've built perception systems for.

Meta Reality Labs

Seattle

Confidential research & development.

CalTech

Pasadena

Fish monitoring next to river turbines in Alaska.

S3 Global

Redmond

Real-time markerless motion capture of professional soccer matches.

Playfair Capital

London

Technical due diligence of investment candidates.

Code19 Racing

Abu Dhabi

Lidar-based opponent racecar detection and tracking for the world's first professional autonomous racing team.

AIPEX Racing

New York City

Computer-vision perception for autonomous racing.

AEDIT

New York City

3D facial scanning as a pre-consultation tool for cosmetic surgery.

AuthenticID

Kirkland

Secure document authentication for identity verification — acquired by Incode.

Incode

San Francisco

Fraud detection for biometric identity verification.

LoomAI

San Francisco

Personal avatar creation from photos — helped launch AR Emoji on Samsung devices.

FitzFrames

Los Angeles

Custom 3D-printed eyewear generated from a 3D face scan.

Blue Ring Imaging

Tampa

Underwater 3D reconstruction.

Big Vision

San Diego

Consulting services in computer vision and AI.

OpenCV.ai

Palo Alto

The for-profit consulting arm of OpenCV.org.

IceCream Labs

San Francisco

Visual search.

Atollogy

San Francisco

Computer vision for manufacturing operations.

Tailored

Naples

Automated clothing measurements from imagery.

Shoebox

New York City

Trading-card grading via automated damage detection.

Univ. of Alaska

Anchorage

Fish monitoring next to river turbines.

Recognition

Research & honors.

  • 2010Best Student Paper, BMVCBritish Machine Vision Conference
  • 2010Foley Scholars AwardGeorgia Tech research competition
  • 2010Ph.D., Computer ScienceGeorgia Institute of Technology
  • ILMScientific & Engineering AwardAcademy of Motion Picture Arts and Sciences — contributed to a 3D scanning system at Industrial Light & Magic
  • Co-founder, PlayMotionGesture-recognition systems deployed across three Disney park divisions

1,100+ citations across computer vision & graphics.

Peer-reviewed work in projector-camera systems, motion-coherent tracking and video synthesis, with publications in IJCV, CVPR, and ACM UIST.

Patents

A portfolio of computer-vision patents.

Inventor on more than ten U.S. patents and applications — spanning gesture control, interactive displays, virtual try-on, active lighting and identity-document security.

Alternative control of touch-based devices

US 9,658,695 · US 9,671,874 · US 10,108,271 · US 11,237,638 · US 12,099,658

Computer-vision gesture control of mobile and touch devices from a distance — operating apps on an external display without ever touching the device.

Calibration & use for an interactive imaging environment

US 8,059,894 · US 8,867,835

Auto-calibrating interactive displays that align human motion to projected imagery, from the PlayMotion era.

Crowd-based visual display with independently-moving pixels

US 8,049,688 · US 2012/0026075

Turning a crowd into a coordinated display, each participant acting as an independently-controlled pixel.

Simulating accessory display on a subject

US 9,058,605 — Taaz

Photorealistic virtual try-on of eyewear and accessories from a single photo.

Interactive imaging systems for motion control

US 8,538,153 · US 9,342,165 · US 9,674,514 · US 10,038,896 · US 11,006,103

Gesture- and motion-driven control of imaging systems, enabling meaningful interaction with projected video.

Digital image generation through an active lighting system

US 11,195,047

Producing enhanced digital imagery using a controlled active-lighting system.

Tamper detection for identification documents

US 10,534,971 · US 2022/0406029 · US 2024/0265721 · WO 2018/071768 — ID Metrics

Detecting forgery and manipulation in ID documents for secure identity verification.

The collective

Collaborators & teammates.

Contract projects are executed by Matt Flagg and a deep bench of longtime collaborators — engineers and researchers across leading CV and AI shops.

Get in touch

Have a hard perception problem?

Tell us what you're building. We take on a focused number of engagements and embed deeply with each one.

mflagg@gmail.com