iustin94 / sign-academy

Sign language recognition using leap motion controller using decision tree model. This hack is made during AUHack2017

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

sign-academy

Sign language alphabet intepreter.

This is a TypeScript framework that uses the Leap motion camera to interpret hand signs into sign language alphabet.The project was developed as our team's entry for the #AUHack2017 event. The framework uses the raw data provided by the camera and interprets the vector data into specific conditions that need to be fulfilled to obtain a specific letter.

For determining the letter symbol we have used the publication of Makiko Funasaka, Yu Ishikawa, Masami Takata, and Kazuki Joe Sign Language Recognition using Leap Motion Controller. Our prliminary KPI mesurments are positive, with accuracy around 67% far away from the 82% discribed in the publication.

To demonstrate the functionality the project has a simple Web UI that tests your ability to spell sign alphabet.

Devpost: link

Correct: alt text

Uncorrect: alt text

About

Sign language recognition using leap motion controller using decision tree model. This hack is made during AUHack2017

License:Apache License 2.0


Languages

Language:CSS 59.5%Language:TypeScript 21.6%Language:JavaScript 10.8%Language:HTML 8.1%