Abstract

The Nine Lives is a mysterious speakeasy that only who find the ticket and solve the bouncer’s riddle can access. Starting from the brief “use the machine learning to create a lock” this prototype uses image recognition to check the user’s input and provide them with the door’s key, but only if they can solve the puzzle like experience offered by the creepy figure guarding the entrance.

Selected images

DSC_9268-Edit.jpg

DSC_9258.jpg

DSC_9249-Edit.jpg

DSC_9288.jpg

DSC_9275-Edit-Edit.jpg

Concept

Inspired by the Unlock brief, the project simulates a mysterious speakeasy accessible only to those who find a special ticket and solve the bouncer's riddle. The concept draws on themes of secrecy, exclusivity, and problem-solving, incorporating puzzles that engage users through audio and visual elements.

The visual identity, prototype, and animations reflect the enigmatic atmosphere of the speakeasy, creating a cohesive and immersive narrative.

The experience begins when the user discovers the Nine Lives ticket, which provides instructions to knock three times on the hatch. Following the knock, the bouncer appears and asks to see the ticket. When the ticket is shown, the bouncer tells a riddle that hints at how to fold the ticket. Only when the ticket is correctly folded, the user gain access to the speakeasy.

Prototyping

The prototyping phase focused on integrating machine learning to enable seamless interaction. Audio recognition systems were used to detect three consecutive knocks on the hatch, ignoring other sounds. Visual recognition systems were trained to identify the ticket and distinguish it from hands or backgrounds.

Advanced training ensured that only tickets folded correctly to form the logo were accepted. Incorrectly folded tickets were recognized and rejected, ensuring the puzzle had a single, clear solution. The physical prototype featured magnetic closures, allowing for easy adjustments and repairs during testing. The hatch design and ticket graphics were crafted to reflect the mysterious and exclusive nature of the speakeasy, reinforcing the overall theme.

Coding

The prototype is developed entirely in JavaScript using Google's Teachable Machine API. The code consists of a state machine that, depending on the passed time and the current stage of the experience, displays different animations, plays distinct sounds, and checks what the camera sees or the microphone hears.