Research / Computer vision · Dataset methodology

LaSalle DB1

Database Augmentation and Independence Testing for Facial Recognition. A dataset-and-methodology study examining augmented data across four recognition approaches.

Accepted for presentation

LaSalle DB1 — original concept placeholder, not gameplay

My role

Primary Author

Team

6 authors

Built with

LBPH · Eigenfaces · Fisherfaces · SFace

Year

2026

01 / MY CONTRIBUTION

The work behind the experience.

I led the construction of LaSalle DB1, developed an augmentation taxonomy, and worked on an independence-testing framework to examine the reliability of augmented data.

02 / OUTCOME

Where it led.

Accepted for presentation at IW-FCV 2026. Scheduled for the October 1 poster session. Proceedings publication is a separate process; no headline benchmark score is claimed.

Andrew Eroyla, Kyle, John Roland, Jim Jonathan, Loreto, and Weon Geun Oh. Full publication credits will be added with the final paper.

Have something in mind?

Let’s create something