Caleb Heinzman
HomeWorkPublicationsPressOpen SourceAbout
← Back to Work
Deeptector logo

Deeptector

Deepfake detection venture delivering a practical browser-based tool for video authenticity analysis. Built end-to-end preprocessing for frame-level analysis, trained MobileNet/Xception/autoencoder/LSTM models, and validated with rigorous benchmarks.

Product Demo

System Architecture & Results

High-level architecture, model overview, and representative evaluation results.

Deeptector architecture diagram
Deeptector model overview
Deeptector results visualization

Press

faviconUM System News: Deepfake-Fighting Tool Wins CompetitionPRESS

UM System coverage of RJI competition win

faviconUMKC News CoveragePRESS

University recognition for deepfake detection platform

faviconStudents Do Business for Social CausesPRESS

University of Missouri research spotlight

faviconImagine Cup 2020 World FinalistsPRESS

Global finalist recognition

faviconDeeptector Demo VideoPRESS

Technical demonstration and overview

Awards

faviconRJI Student Competition WinnerAWARD

Reynolds Journalism Institute innovation award

faviconMicrosoft Imagine Cup — Americas Winner, World FinalistAWARD

Americas Regional Winner; advanced to World Finals

Caleb Heinzman

Building reliable machine learning systems for healthcare, finance, and autonomy. Research-minded and product-driven.

Navigation

WorkPublicationsPressOpen SourceAbout

Connect

GitHubLinkedIn

© 2025 Caleb Heinzman. All rights reserved.

Built with Next.js, TypeScript, and TailwindCSS