A Generic Multi-modal Dynamic Gesture Recognition System Using Machine Learning
Human computer interaction facilitates intelligent communication between humans and computers, in which gesture recognition plays a prominent role. This paper proposes a machine learning system to identify dynamic gestures using tri-axial acceleration dat
- PDF / 1,005,986 Bytes
- 13 Pages / 439.37 x 666.142 pts Page_size
- 26 Downloads / 191 Views
Research and Outreach, Solarillion Foundation, Chennai, India 2 College of Engineering, Guindy, Chennai, India 3 SRM University, Chennai, India {gautham.krishna,nathankarthik,yogesh.bkumar,ankithprabhu, ajaykannan,vineethv}@ieee.org
Abstract. Human computer interaction facilitates intelligent communication between humans and computers, in which gesture recognition plays a prominent role. This paper proposes a machine learning system to identify dynamic gestures using tri-axial acceleration data acquired from two public datasets. These datasets, uWave and Sony, were acquired using accelerometers embedded in Wii remotes and smartwatches, respectively. A dynamic gesture signed by the user is characterized by a generic set of features extracted across time and frequency domains. The system was analyzed from an end-user perspective and was modelled to operate in three modes. The modes of operation determine the subsets of data to be used for training and testing the system. From an initial set of seven classifiers, three were chosen to evaluate each dataset across all modes rendering the system towards mode-neutrality and dataset-independence. The proposed system is able to classify gestures performed at varying speeds with minimum preprocessing, making it computationally efficient. Moreover, this system was found to run on a low-cost embedded platform – Raspberry Pi Zero (USD 5), making it economically viable. Keywords: Gesture recognition · Accelerometers Feature extraction · Machine learning algorithms
1
Introduction
Gesture recognition can be defined as the perception of non-verbal communication through an interface that identifies gestures using mathematical, probabilistic and statistical methods. The field of gesture recognition has been experiencing a rapid growth amidst increased interests shown by researchers in the industry. The goal of current research has been the quick and accurate classification of gestures with minimalistic computation, whilst being economically feasible. c Springer Nature Switzerland AG 2019 K. Arai et al. (Eds.): FICC 2018, AISC 887, pp. 603–615, 2019. https://doi.org/10.1007/978-3-030-03405-4_42
604
G. G. Krishna et al.
Gesture recognition can find use in various tasks such as developing aids for the audio-vocally impaired using sign language interpretation, virtual gaming and smart home environments. Modern gesture recognition systems can be divided into two broad categories vision based and motion based systems. The vision based system proposed by Chen et al. in [1] uses digital cameras, and that proposed by Biswas et al. in [2] uses infrared cameras to track the movement of the user. For accurate classification of the gestures, these systems require proper lighting, delicate and expensive hardware and computationally intensive algorithms. On the other hand, motion based systems use data acquired from sensors like accelerometer, gyroscope and flex sensor to identify the gestures being performed by the user. Of late, most gesture recognition systems designed for effective interaction u
Data Loading...
