3DCNN+LSTM Gesture Recognition for Dashboard Interaction

Ong, Zi Yuan (2022) 3DCNN+LSTM Gesture Recognition for Dashboard Interaction. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Ong Zi Yuan_Fulltext.pdf Restricted to Registered users only Download (3MB)

Abstract

In today’s world of continuously advancing computing technology, recognition technology is one such example that has managed to make its way into the mainstream. One growing application of recognition technology is gesture recognition, which has a lot of potential uses, in many fields and products. This includes gestures to control consumer devices like phones, gestures used to control robotics and wheelchairs, to even recognizing and translating sign languages. Gesture recognition initially was a niche field, but has now grown in relevance thanks to recent advancement. However, there is a lack of applications of gesture recognition in daily life, especially in the usage of consumer products. The project aims to develop a real-time hand gesture recognition system that can be used to interact with a dashboard on a PC, by researching the various gesture recognition techniques and approaches and evaluating their performance and effectiveness. The gesture recognition system is developed to interact with Spotify’s User Interface on a Personal Computer, which receives input from the webcam of a computer, detect the gestures of the user, and then translate those gestures to actions on Spotify’s dashboard. These actions will be used to control Spotify, such as to skip to the next track, or to the previous track, or to change volume, and also to control devices inside AIoT Smart Classroom such as aircon, lights and door.