Scholars Central

Conference abstractPublished 25 Sep 2017

Real-time gesture recognition using dynamic time warping and simplification methods

Authors: Alan Soares

Federal University of Bahia, Brazil

Ask about this article

Abstract

Select any words in the title or abstract, or press this and choose a sentence.

The recognition of dynamic gestures of hands using pure geometric 3D data in real-time is a challenge. RGB-D sensors simplified this task, giving an easy way to acquire 3D points and track them using the depth maps information. But use this collection of raw 3D points as a gesture representation in a classification process is prone to mismatches, since gestures of different people can vary in scale, location and velocity. In this paper we analyze how different techniques of simplification and regularization can provide more accurate representations of the gestures. Using Dynamic Time Warping (DTW) as the classification method, we show that the simplification and regularization steps can improve the recognition rate and also reduce the time of gesture recognition.

Ask about this article

Answers come only from this article's title, abstract and keywords, each sentence pointing to where it comes from. Your question is sent to OpenAI to write the answer. Scholars Central does not store it.

Try:

For authors and organisers

Publish, present and keep up with your field

Find the right journal for your paper and submit your manuscript.

Photo: Pexels