Title: Hand Gesture Recognition using Double CNN and Transfer Learning
Abstract: Hand gesture recognition forms a very integral part of human-computer interaction (HCI). With the advent of augmented reality, it becomes even more important for a computer system to understand visual cues obtained from the environment. While easier to design the system with the usage of physical sensors placed on the user hands, it is not an economically feasible. The aim of this work is to understand and implement various hand gesture recognition models to work on webcams and mobile cameras. The implementations will target both static gestures and dynamic gestures. For the former, an image or a single frame can be used to determine the gesture. For the latter, a series of images, or a video clip is used to determine a gesture. For the detection of the gesture, we employ Double-channel CNN over a smaller part of the dataset to detect the presence of the gesture. Once a gesture is detected we use the skeletal based approach along with transfer learning to classify the gesture.
Publication Year: 2021
Publication Date: 2021-06-16
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot