Home Automation Oriented Gesture Classification From Inertial Measurements

Abstract

In this paper, a machine learning (ML) approach is presented that exploits accelerometers data to deal with gesture recognition (GR) problems. The proposed methodology aims at providing high accuracy classification for home automation systems, which are generally user independent, device independent, and device orientation independent, an heterogeneous scenario that has not been fully investigated in previous GR literature. The approach illustrated in this paper is composed of three main steps: event identification; feature extraction; and ML-based classification. The elements of the novelty of the proposed approach are 1) a preprocessing phase based on principal component analysis to increase the performance in real-world scenario conditions and 2) the development of parsimonious novel classification techniques based on sparse Bayesian learning. This methodology is tested on two datasets of four gesture classes (horizontal, vertical, circles, and eight-shaped movements) and on a further dataset with eight classes. In order to authentically describe a real-world home automation environment, the gesture movements are collected from more than 30 people who freely perform any gesture. It results in a dictionary of 12 and 20 different movements, respectively, in the case of the four-class and the eight-class databases

Design and evaluation of novel gesture-centric interaction models in smart spaces

“…Based on the revision of the user studies on healthy smart homes, Kim et al [45] listed several research issues, including that i) the systems should be developed based on the needs of the end users instead of on technological capabilities, that ii) it is essential that intuitive user interfaces are designed, that iii) the end users' feedback about how the system performances in actual use is essential, that iv) user experience should be considered for a successful technology interaction and adoption of health smart systems, and that v) the consideration of the usability of the system is essential to increase the acceptability of the system. Cenedese et al [18] also emphasize the importance to consider the man-in-theloop issue and to pose the centrality of human behavior in terms of context awareness and interface.…”

“…Most of the works about gesture recognition approaches in the review literature are focused on gesture data from optical [107,31,96] or inertial sensor [52,18,37], or the fusion of these two [21,65,20]. This work, however, explores a unified solution to the gesture recognition problem based on both types of sensors, widening the base of compatible gesture input devices in terms of sensing devices.…”

“…Although the vision-based tracking enables the users to do free gestures without cumbersome contact-based devices like data glove, it does have some limitations, such as being prone to be interfered by varying lighting conditions and cluttered background and relatively low sampling rate of normal cameras [21]. The inertial based recognition does not have the [18,109] utilize accelerometer to record the gesture motions while others combine accelerometer with gyroscope [5,54,108,52] and magnetometer [37]. However, the inertial based gesture recognition presents the drawback that the tracking is implicit and it cannot collect position and orientation information.…”

“…Its collected database provides a common ground to compare recognition performance given different tracking signals and approaches. Sometimes the features to extract are obvious and highly related to the classifier such as the case of [18] where the two RVM (Relevance Vector Machine) approaches requires the calculations of frame-based descriptor and DTW matching. Some features are specifically designed to better represent the characteristics of gestures, to fuse different data types, to cope with the variances of gesture performance and captured signal, or to be indexed for recognition:…”

“…In addition, the orientation features indicating the rotation state of a trajectory are complement of the shape features and help distinguish gestures with similar shapes. The approach in [18] is enhanced and device-orientation-independent with a proposed tool called PrePCA. Sign sequences of acceleration data are extracted to compress data and to minimize the influence of variations between users in [109].…”