Dynamic time warping

About: Dynamic time warping is a research topic. Over the lifetime, 6013 publications have been published within this topic receiving 133130 citations.

Inferring Directed Road Networks from GPS Traces by Track Alignment

Abstract: This paper proposes a method to infer road networks from GPS traces These networks include intersections between roads, the connectivity between the intersections and the possible traffic directions between directly-connected intersections These intersections are localized by detecting and clustering turning points, which are locations where the moving direction changes on GPS traces We infer the structure of road networks by segmenting all of the GPS traces to identify these intersections We can then form both a connectivity matrix of the intersections and a small representative GPS track for each road segment The road segment between each pair of directly-connected intersections is represented using a series of geographical locations, which are averaged from all of the tracks on this road segment by aligning them using the dynamic time warping (DTW) algorithm Our contribution is two-fold First, we detect potential intersections by clustering the turning points on the GPS traces Second, we infer the geometry of the road segments between intersections by aligning GPS tracks point by point using a “stretch and then compress” strategy based on the DTW algorithm This approach not only allows road estimation by averaging the aligned tracks, but also a deeper statistical analysis based on the individual track’s time alignment, for example the variance of speed along a road segment

Hand Movements and Gestures Characterization Using Quaternion Dynamic Time Warping Technique

Abstract: Trends show that wearable devices with high-range sensors offer new business opportunities in enriching user experience in swing-based outdoor games, such as tennis, golf, and/or indoor gaming applications. Such applications providing insights into a player’s abilities to play these games require methods that efficiently distinguish and capture intricacies of hand movements and/or gestures. In this paper, we show that quaternions-based dynamic time warping (QDTW) technique provides an efficient means for characterizing different arm/hand movements and gestures. A complete methodology and results for the pursued case study of outdoor tennis game are provided in this paper. We propose a new and unique approach for training data for various tennis shots and then using DTW and QDTW at the two levels of a hierarchical classifier for classification of an incoming tennis shot. The achieved accuracy for tennis shots detection is more than 99% and that for classification is 90%. Furthermore, the concept of consistency in a player’s shots and how a played shot differs from a professional’s similar shot are considered to suggest recommendations for improvement to the player.

Toward Force-Based Signature Verification: A Pen-Type Sensor and Preliminary Validation

Abstract: A compact pen-type force sensor is developed to study the feasibility of force-based signature verification. A force-sensing method based on leverage effect is proposed to detect 3-D forces between the pen's tip and the paper. A compact and low-cost force-sensing assembly is designed, which is integrated by five off-the-shelf 1-D force sensors. A matrix-based measurement model is established to compute the force signal in the task coordinate system (CS), which is transformed from the force signal in the sensor CS and the angle signal from a 2-D angle sensor. The structural parameters of the force sensor are determined both to achieve the required force accuracy and to meet the constraints of pen size for comfortable grasping. System performance experiments are carried out to measure the absolute and repetitive accuracy of the pen. The results show that the pen is capable for detection of 3-D force signals during real-time handwriting. Repetitive accuracy is measured to be about 0.05 N. Finally, a small-scale signature verification experiment is carried out. The verification results based on the dynamic time warping (DTW) method show that the equal error rate (EER) is about 6.3%, which illustrates the potential of the pen for force-based signature verification.