Gesture Recognition for Harry Potter attraction at Universal Parks

Universal Studios has 5 theme parks across the US, Japan, Singapore and China, with a daily attendance of 100.000+ visitors.

Inaugurated in 2010, The Wizarding World of Harry Potter (present in 4 of these parks) rapidly became one of its main attractions. Thousands of visitors and "Potterheads" (Harry Potter's fandom) travel from remote corners of the world to be enchanted by their favorite characters and magical surroundings.

Source: Universal Parks

CHALLENGE

Among the activities offered to visitors is the experience of casting spells in Diagon Alley, the apprentice magicians' market. These spells, cast by magic wands, trigger interactions with various elements in the surroundings.

The challenge was to allow guests to activate interactive props when performing "spells". This implied accurately tracking the wands’ tips in the unconstrained environment of the park.

The solution required recognizing gestures with a high degree of precision and reliability in order to provide guests with the experience they expect.

SOLUTION

We designed a Near-Infrared Video Acquisition System paired with a suitable IR LED (infrared light emitting diode) installed in the interactive wands' tips.

This solution enabled us to perform accurate wand tracking in both daylight and darkness.

We also developed two complementary gesture recognition models to recognize the spells with high accuracy.

This project was the basis for the interactive attraction in production to this day in the US, China and Japan. Universal relied on Digital Sense's experts to develop unprecedented technology that continues to amaze millions of visitors each year.

Our work

See more cases

Ulta Beauty Case Study - Make-up try-on experience with augmented reality

User experience-focused solution for Ulta Beauty, the largest beauty retailer in the USA. An augmented reality makeup try-on development that provides a realistic real-time experience.

Bader Case Study - Automatic defect detection in premium leather enterprise

Quality control station development through deep-learning models that detect potential defects with very high accuracy and pinpoint them for the technician.

"We came up with a problem, a rough idea on how to solve it, and worked out with them a technical solution. Most of our projects are R&D focused and complex on emerging tech. They augmented our team and established a very good relationship. They deliver, they put forward ideas and I like that they are willing to stand their ground for the good of the project instead of just trying to be agreeable."
Marc Fisher
Dogtown Media, CEO
"We have a great partnership with Digital Sense. They enhance and complement our computer vision R&D efforts. We enjoy working with the team and we are looking forward to continuing to grow our partnership!"
Agustina Sartori
Ulta Beauty, Director of AR Innovation
"In 2018 we partnered with them to help us develop a custom OCR module for our iOS and Android mobile applications. Their team did a great job of scoping out the project and developing the OCR for our app's use case. We continue to work with them today on improvements and leveraging the latest machine learning capabilities to deliver optimal performance to our app users. We highly recommend Digital Sense."

John Gushman
Prepaid2cash, Co-Founder

Contact

Get in touch

Contact us
marketing@digitalsense.ai
US: +1 415 766 0271
UY: +598 2 614 1674

Resources
Blog
Portfolio
Careers

Connect