Dynamic Time Warping Technique Research Articles

An accurate slicing method for dynamic time warping algorithm and the segment-level early abandoning optimization

Time series data analysis algorithms have been gaining significant importance in the research community. Extensive studies have confirmed that Dynamic Time Warping (DTW) is the best distance measure in time series analysis across multiple domains. However, DTW is a time-consuming algorithm with quadratic time complexity, which limits its widespread adoption. In this paper, we proposed a novel slicing mechanism for DTW, called Partial Dynamic Time Warping (PDTW). PDTW is capable of dividing a complete DTW calculation into multiple independent partial calculations. The proposed PDTW ensures very consistent alignments with the original Constrained DTW (CDTW) by incorporating additional data windows for each pairwise time series. On the basis of PDTW, we also present Early Abandoning Dynamic Time Warping (EADTW) technique, which saves a substantial amount of computing time by abandoning superfluous calculations of unnecessary segments in the time series classification task. Large-scale experimental results on 96 datasets from the UCR archive show the effectiveness of PDTW and EADTW. The cumulative sum of PDTW distances for segments is nearly identical to the CDTW distance, and in many cases, it is exactly equal. This remarkable characteristic of PDTW indicates its immense potential for further development and optimization of DTW, such as parallelization. In classification tasks, EADTW is 1.99 times faster than CDTW on average, with a maximum speedup of 2.89 times, while maintaining accuracy. Additionally, it enhances accuracy by 0.93% and improves speed by 2.34 times under the accuracy priority parameter.

Time Series Clustering Analysis Using Dynamic Time Warping Technique of Daily Rainfall in Bengkulu Province

Time Series Clustering Analysis Using Dynamic Time Warping Technique of Daily Rainfall in Bengkulu Province

Employing the Coupled EUHFORIA‐OpenGGCM Model to Predict CME Geoeffectiveness

AbstractEUropean Heliospheric FORecasting Information Asset (EUHFORIA) is a physics‐based data‐driven solar wind and coronal mass ejections (CMEs) propagation model designed for space weather forecasting and event analysis investigations. Although EUHFORIA can predict the solar wind plasma and magnetic field properties at Earth, it is not equipped to quantify the geo‐effectiveness of the solar transients in terms of geomagnetic indices like the disturbance storm time (Dst) index and the auroral indices, that quantify the impact of the magnetized plasma encounters on Earth's magnetosphere. Therefore, we couple EUHFORIA with the Open Geospace General Circulation Model (OpenGGCM), a magnetohydrodynamic model of the response of Earth's magnetosphere, ionosphere, and thermosphere to transient solar wind characteristics. In this coupling, OpenGGCM is driven by the solar wind and interplanetary magnetic field obtained from EUHFORIA simulations to produce the magnetospheric and ionospheric response to the CMEs. This coupling is validated with two observed geo‐effective CME events driven with the spheromak flux‐rope CME model. We compare these simulation results with the indices obtained from OpenGGCM simulations driven by the measured solar wind data from spacecraft. We further employ the dynamic time warping (DTW) technique to assess the model performance in predicting Dst. The main highlight of this study is to use EUHFORIA simulated time series to predict the Dst and auroral indices 1–2 days in advance, as compared to using the observed solar wind data at L1, which only provides predictions 1–2 hr before the actual impact.

Unsupervised neural decoding for concurrent and continuous multi-finger force prediction

Reliable prediction of multi-finger forces is crucial for neural-machine interfaces. Various neural decoding methods have progressed substantially for accurate motor output predictions. However, most neural decoding methods are performed in a supervised manner, i.e., the finger forces are needed for model training, which may not be suitable in certain contexts, especially in scenarios involving individuals with an arm amputation. To address this issue, we developed an unsupervised neural decoding approach to predict multi-finger forces using spinal motoneuron firing information. We acquired high-density surface electromyogram (sEMG) signals of the finger extensor muscle when subjects performed single-finger and multi-finger tasks of isometric extensions. We first extracted motor units (MUs) from sEMG signals of the single-finger tasks. Because of inevitable finger muscle co-activation, MUs controlling the non-targeted fingers can also be recruited. To ensure an accurate finger force prediction, these MUs need to be teased out. To this end, we clustered the decomposed MUs based on inter-MU distances measured by the dynamic time warping technique, and we then labeled the MUs using the mean firing rate or the firing rate phase amplitude. We merged the clustered MUs related to the same target finger and assigned weights based on the consistency of the MUs being retained. As a result, compared with the supervised neural decoding approach and the conventional sEMG amplitude approach, our new approach can achieve a higher R2 (0.77 ± 0.036 vs. 0.71 ± 0.11 vs. 0.61 ± 0.09) and a lower root mean square error (5.16 ± 0.58 %MVC vs. 5.88 ± 1.34 %MVC vs. 7.56 ± 1.60 %MVC). Our findings can pave the way for the development of accurate and robust neural-machine interfaces, which can significantly enhance the experience during human-robotic hand interactions in diverse contexts.

A novel approach for residential load identification based on dynamic time warping

This paper proposes a new approach for load identification using a combination of fast Fourier transform (FFT) and dynamic time warping (DTW) techniques. In this approach, first, the current signals of each load device are measured individually and in various combinations to generate reference signals. FFT is then used to obtain meaningful features from the measured signals, while DTW is employed to classify the loads. This is the first known study to use the DTW algorithm for residential load identification in this structure. The proposed approach is constructed in two ways: supervised (DTW-SUP) and semi-supervised (DTW-SEM). In DTW-SEM, the reference signals for the load combinations are generated by summing the sample measurements of the individual loads, while in DTW-SUP, the actual sample measurements are used. The approach has been tested on six loads, including different load combinations. It has achieved high accuracy levels of over 90% in identifying single loads and load combinations using the error metrics Precision, Recall, and F-score. The proposed approach can potentially facilitate the implementation of energy-saving strategies in residential buildings and improve energy efficiency.

Micro-short circuit fault diagnosis of lithium-ion battery based on voltage curve similarity ranking volatility

ABSTRACT An electric vehicle battery system often consists of multiple battery cells that are interconnected in series and parallel. If a micro-short circuit failure occurs in one of the battery cells, it may cause a serious internal short circuit or even thermal runaway. Consequently, it is extremely important to detect micro-short circuit faults in batteries. This paper presents a novel approach for diagnosing faults in lithium-ion batteries based on the similarity ranking fluctuation rate of voltage curve, and verify the feasibility of the method through a series of micro-short circuit experiments containing an external parallel variable resistance device. This approach does not require complete charge-discharge curves. Firstly, the charging voltage data is preprocessed using the Variational Mode Decomposition algorithm, after which the Dynamic Time Warping technique is employed to calculate the similarity between the charging voltage of each battery and a reference battery. Then a ranking fluctuation rate of voltage curve similarity is established, and the Grubbs criterion is utilized to identify batteries exhibiting abnormal fluctuation rate within the ranking. Finally, the effectiveness of the proposed method is validated by analyzing actual battery signals collected from faulty vehicles. The results demonstrate that the method accurately identifies batteries with micro-short circuit faults.

Time Series Classification with Multiple Wavelength Scattering Signals for Nuisance Alarm Mitigation

Smoke detectors are the most widely used fire detectors due to their high sensitivity. However, they have persistently faced issues with false alarms, known as nuisance alarms, as they cannot distinguish smoke particles, and their responsiveness varies depending on the particle size and concentration. Although technologies for distinguishing smoke particles have shown promising results, the hardware limitations of smoke detectors necessitate an intelligent approach to analyze scattering signals of various wavelengths and their temporal changes. In this paper, we propose a pipeline that can distinguish smoke particles based on scattering signals of various wavelengths as input. In the data extraction phase, we propose methods for extracting datasets from time series data. We propose a method that combines traditional approaches, early detection methods, and a Dynamic Time Warping technique that utilizes only the shape of the signal without preprocessing. In the learning model and classification phase, we present a method to select and compare various architectures and hyperparameters to create a model that achieves the best classification performance for time series data. We create datasets for six different targets in our presented sensor and smoke particle test environment and train classification models. Through performance comparisons, we identify architecture and parameter combinations that achieve up to 98.7% accuracy. Ablation studies under various conditions demonstrate the validity of the chosen architecture and the potential of other models.

Interpersonal coordination analysis in bat-and-ball sports under a real game situation: Asymmetric interaction and delayed coupling

This study investigated the interpersonal coordination between the pitcher and the batter in bat-and-ball sports. Although the importance of interpersonal coordination is widely accepted in many sports, no studies have investigated it in bat-and-ball sports because the dominant task constraints surrounding the interaction between pitcher and batter make it difficult to apply conventional analytic techniques. To address the issue, this study proposes a new analytical framework to investigate interpersonal coordination in bat-and-ball sports under a real game situation with two main characteristics: asymmetric interaction and delayed coupling. First, the dynamic time warping technique was used to evaluate the stability of the head movement pattern of the pitcher and batter, and cross-correlation analysis was used to quantify the temporal relationship between them. We found that the head movement pattern of batters was significantly more unstable than that of pitchers, and approximately 60% of the variance of the change in the head movement pattern of batters could be explained by that of the pitchers. Moreover, expert batters followed a pitcher’s movements with a specific time delay of approximately 250 ms. These findings highlight the characteristics of interpersonal coordination in bat-and-ball sports: the pitcher can make a pre-patterned stable motion, whereas the batter needs to follow and adjust their movement to it. Although the effects of prediction ability need to be investigated to understand its detailed mechanism, the contribution of this study is that it revealed the existence of the interpersonal coordination between the pitcher and batter of bat-and-ball sports under a real game situation.

Assessment of STI dry etch process variability by means of dynamic time warping technique

Numerical data of two in situ optical acquisition systems were used in machine learning algorithm to evaluate an shallow trench isolation dry etch process in a dataset of more than 200 etched wafers processed during a year. Though overall recipe performance was according to specifications, the observed parameter fluctuations were characterized to check for correlation patterns in processing data by means of machine learning. Thus, dynamic time wrapping was used to analyze time series datasets to get characteristics of minimum distance’s path between signals coming of each individual wafer. Such metrics of distances could help clustering groups of wafers that appear to be in proximity. Out of eight analyzed channels, we observed the largest distance variability was in the Si main etch for spectral reflectometer dataset, which we compared to optical emission spectroscopy channels. Based on correlation assessment of distances to the etch depth uniformity data, we showed that the selected approach is a good means to analyze time series datasets of e.g. dry etch processes for monitoring process stability.

TIMIT-TTS: A Text-to-Speech Dataset for Multimodal Synthetic Media Detection

With the rapid development of deep learning techniques, the generation and counterfeiting of multimedia material has become increasingly simple. Current technology enables the creation of videos where both the visual and audio contents are falsified. While the multimedia forensics community has begun to address this threat by developing fake media detectors. However, the vast majority existing forensic techniques only analyze one modality at a time. This is an important limitation when authenticating manipulated videos, because sophisticated forgeries may be difficult to detect without exploiting cross-modal inconsistencies (e.g., across the audio and visual tracks). One important reason for the lack of multimodal detectors is a similar lack of research datasets containing multimodal forgeries. Existing datasets typically contain only one falsified modality, such as deepfaked videos with authentic audio tracks, or synthetic audio with no associated video. Currently, datasets are needed that can be used to develop, train, and test these forensic algorithms. In this paper, we propose a new audio-visual deepfake dataset containing multimodal video forgeries. We present a general pipeline for synthesizing deepfake speech content from a given video, facilitating the creation of counterfeit multimodal material. The proposed method uses Text-to-Speech (TTS) and Dynamic Time Warping (DTW) techniques to achieve realistic speech tracks. We use this pipeline to generate and release TIMIT-TTS, a synthetic speech dataset containing the most cutting-edge methods in the TTS field. This can be used as a standalone audio dataset, or combined with DeepfakeTIMIT and VidTIMIT video datasets to perform multimodal research. Finally, we present numerous experiments to benchmark the proposed dataset in both monomodal (i.e., audio) and multimodal (i.e., audio and video) conditions. This highlights the need for multimodal forensic detectors and more multimodal deepfake data.

Driving Maneuvers Recognition and Classification Using A Hyprid Pattern Matching and Machine Learning

Since most of the road and traffic accidents are related to human errors or distraction, the study of irregular driving behaviors is considered one of the most important research topics in this field. To prevent road accidents and assess driving competencies, there is an urgent need to evaluate driving behavior through the design of a driving maneuvers assessment system. In this study, the recognition and classification of highway driving maneuvers using smartphones’ build-in sensors are presented. The paper examines the performance of three classical machine learning techniques and a novel hybrid system. The proposed hybrid system combines the pattern machining Dynamic Time Warping (DTW) technique for recognizing driving maneuvers and the machine learning techniques for classification. Results obtained from both approaches show that the performance of the hybrid system is superior to that obtained by using classical machine learning techniques. This enhancement in the performance of the hybrid system is due to the elimination of the overlapping in the target classes due to the separation, the recognition and the classification processes.

A Passive Online Technique for Learning Hybrid Automata from Input/Output Traces

Specification synthesis is the process of deriving a model from the input-output traces of a system. It is used extensively in test design, reverse engineering, and system identification. One type of the resulting artifact of this process for cyber-physical systems is hybrid automata. They are intuitive, precise, tool independent, and at a high level of abstraction, and can model systems with both discrete and continuous variables. In this article, we propose a new technique for synthesizing hybrid automaton from the input-output traces of a non-linear cyber-physical system. Similarity detection in non-linear behaviors is the main challenge for extracting such models. We address this problem by utilizing the Dynamic Time Warping technique. Our approach is passive, meaning that it does not need interaction with the system during automata synthesis from the logged traces; and online, which means that each input/output trace is used only once in the procedure. In other words, each new trace can be used to improve the already synthesized automaton. We evaluated our algorithm in one industrial and two simulated case studies. The accuracy of the derived automata shows promising results.

On analyzing GNSS displacement field variability of Taiwan: Hierarchical Agglomerative Clustering based on Dynamic Time Warping technique

We investigate the feasibility of using the Dynamic Time Warping (DTW) technique to cluster continuous GNSS displacements in Taiwan. Using the DTW distance as the measure for waveform similarity, we combine the DTW method with the Hierarchical Agglomerative Clustering (HAC) algorithm. This is in contrast to the conventional clustering approach that uses the Euclidean distance, considering the average long-term crustal motion, but inherently neglects full-waveform temporal variations. Here we apply the DTW-based HAC algorithm adopting DTW distance as the waveform similarity measure on 11 years worth of 3-D displacement data from 115 continuous GNSS network stations in Taiwan. We demonstrate the efficacy of the DTW-based HAC method in distinguishing the GNSS spatiotemporal variabilities that are consistent with the known, complex tectonic behavior of the region. An open-source Python package has been developed and made available to perform the HAC analysis.

Matching Model of Dance Movements and Music Rhythm Features Using Human Posture Estimation.

The essential issue in music understanding and dance synthesis research is how to improve the degree of matching between dance motions and music rhythm elements. The matching model of dance movements and music rhythm features is created in this study based on human posture estimation research to tackle the problems that existing matching methods are prone to rapid changes in movements and cannot keep the original movement features. The rhythm properties of movement and music data are first analyzed, and then the degree of matching between movement and music pieces is measured using a dynamic time warping technique. A constraint-based dynamic planning approach is also used to synthesize the dance action sequence that best matches the supplied music. Experiments reveal that this model has a matching accuracy of 95.1 percent, which is higher than the two comparative models' matching accuracy of 5.2 percent and 7.6 percent, respectively. The matching score of this paper is high, reaching around 94 percent, according to the user research results. The method suggested in this work has apparent advantages in that it may efficiently help users in the arrangement of desired dance moves or backing music, and it can be applied in the field of actual choreography and scoring.

Dysarthric Speech Enhancement Based on Convolution Neural Network.

Generally, those patients with dysarthria utter a distorted sound and the restrained intelligibility of a speech for both human and machine. To enhance the intelligibility of dysarthric speech, we applied a deep learning-based speech enhancement (SE) system in this task. Conventional SE approaches are used for shrinking noise components from the noise-corrupted input, and thus improve the sound quality and intelligibility simultaneously. In this study, we are focusing on reconstructing the severely distorted signal from the dysarthric speech for improving intelligibility. The proposed SE system prepares a convolutional neural network (CNN) model in the training phase, which is then used to process the dysarthric speech in the testing phase. During training, paired dysarthric-normal speech utterances are required. We adopt a dynamic time warping technique to align the dysarthric-normal utter-ances. The gained training data are used to train a CNN - based SE model. The proposed SE system is evaluated on the Google automatic speech recognition (ASR) system and a subjective listening test. The results showed that the proposed method could notably enhance the recognition performance for more than 10% in each of ASR and human recognitions from the unprocessed dysarthric speech. Clinical Relevance- This study enhances the intelligibility and ASR accuracy from a dysarthria speech to more than 10.

Modeling of Intelligent Sensor Duty Cycling for Smart Home Automation

The advancement of wireless sensor networks (WSNs) improves various smart home automation services and home users’ living standards. However, efficiently collecting data and automating smart home services require the extensive deployment of the sensors. Thus, one of the crucial and challenging tasks is to minimize the sensors’ energy consumption for monitoring and automating various activities in a smart home. In this article, we present a solution to control the excessive energy consumption of sensors used to detect various activities of daily living (ADL) of a smart home resident. The sensors within a smart home network are divided into various groups employing the recurrent neural network (RNN) and dynamic time warping (DTW) techniques to predict the activities with high accuracy and less energy consumption. The smart home users’ future activities are forecast with bidirectional long short-term memory (BLSTM) RNN model to select those sensors that are likely to predict the upcoming activities. Similarly, to predict the home users’ unusual activities, a guard sensor is elected among sensors with high similarities with each other using DTW. The sensor’s role is evenly switched between different modes to maintain a fair tradeoff between energy and accuracy. An extensive set of simulations is performed to validate the proposed scheme’s work integrating datasets from authentic sources. Finally, the proposed system significantly reduces the sensors’ energy consumption and prolongs the battery lifetime to approximately 137 days. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —This article presented an energy-efficient duty-cycling scheme for automating sensors’ operations in a smart home scenario. The traditional duty-cycling schemes mainly provide solutions based on appointing sentries or predicting smart home users’ next activity using models, such as Bayesian networks. We design a system that integrates the advantages of both sentry and prediction-based schemes to reduce the amount of energy required by sensors to detect and automate smart home users’ activities with high accuracy and precision. The active sensors are appointed using a bidirectional long short-term memory recurrent neural network. Similarly, the guard sensors are assigned to detect unusual activities using the similarities among idle sensors. This study could be used to automate the smart home sensors for detecting home user’s activities with less energy, which ultimately prolongs the battery lifetime of the sensors.

Dynamic Time Warping as an Alternative to Windowed Cross Correlation in Seismological Applications

Abstract We investigate the feasibility of using the dynamic time warping (DTW) technique as an alternative to windowed cross correlation (WCC) for an indirect measure to quantify both the similarity and relationship between two seismic time series. We first examine the sensitivity and performance of the DTW technique by analyzing both synthetic and real seismic time series in geophysical applications. Results show that DTW efficiently retrieves useful information from seismic data and has a high sensitivity to minor variations in time series that WCC fails to detect. We further propose four potential applications of DTW to routine seismic data interpretation—earthquake detection, template matching, clustering of waveforms, and full-waveform inversion for 1D velocity models. The earthquake detection scheme employing DTW is proposed in this study as an alternative to traditional methods. The accuracy of DTW in estimating similarity is explored for template matching and clustering of seismic traces. Finally, we discuss a realistic example of 1D Earth velocity model inversion using DTW and explore its feasibility in full-waveform inversion.

Багатоетапний алгоритм з DTW та кластеризацією для прогнозування середньої депозитної ставки в Україні

The paper is dedicated to the development of the multi-stage forecasting method that is based on Dynamic Time Warping, Clustering and AutoARIMA techniques, which is compared with several traditional benchmarks on the unique dataset. The goal is to forecast an average deposit rate in Ukraine using data that has been scrapped from banks' websites about their individual deposit rates on the daily basis. From this rich dataset the paper focuses only on 12-month deposits, UAH, for each bank. Most of the issues that are traditional for web-scraping approach are irrelevant in our case due to the dataset features. These rates are aggregated into groups by similarity in dynamics, forecasted separately with an AutoARIMA routine and finally aggregated into the entire forecast using weights that have been obtained with an OLS estimation. The paper presents the result and comparison with several benchmarks, starting from simple Random Walk, a few specifications of ARIMA and simple Random Forest. The multi-stage approach outperforms benchmarks by an RMSE and graphical analysis over the latter period of the data.

Grasshopper optimization algorithm based improved variational mode decomposition technique for muscle artifact removal in ECG using dynamic time warping

The removal of unwanted noise from electrocardiogram (ECG) recordings is a difficult procedure in biomedical signal analysis. It alters the signal and affects the accurate interpretation of the signal. A major distortion produced while recording is due to muscle artifact (MA) or electromyographic (EMG) noise. To achieve a high-quality ECG recording by eliminating this MA noise, our research proposes a novel filtering technique using the grasshopper optimization algorithm (GOA) based variational mode decomposition (VMD) method with the dynamic time warping (DTW) distance concept. GOA is utilized to identify the best optimal parameters for the VMD process, and then the DTW technique is applied to discover the relevant modes that include MA noise. To successfully remove noise, these modes are denoised using the discrete wavelet transform (DWT) method. ECG signal analysis is performed on a real-time MIT-BIH arrhythmia database and is compared with the performance of recent existing techniques using signal characteristics such as signal-to-noise ratio (SNR), mean square error (MSE), correlation coefficient (CC), and so on. As an outcome of these computations, our proposed technique outperforms those existing methods when it comes to denoising MA noise.

A Novel Acceleration Signal Processing Procedure for Cycling Safety Assessment.

With the growing rate of urban population and transport congestion, it is important for a city to have bike riding as an attractive travel choice but one of its biggest barriers for people is the perceived lack of safety. To improve the safety of urban cycling, identification of high-risk location and routes are major obstacles for safety countermeasures. Risk assessment is performed by crash data analysis, but the lack of data makes that approach less effective when applied to cyclist safety. Furthermore, the availability of data collected with the modern technologies opens the way to different approaches. This research aim is to analyse data needs and capability to identify critical cycling safety events for urban context where bicyclist behaviour can be recorded with different equipment and bicycle used as a probe vehicle to collect data. More specifically, three different sampling frequencies have been investigated to define the minimum one able to detect and recognize hard breaking. In details, a novel signal processing procedure has been proposed to correctly deal with speed and acceleration signals. Besides common signal filtering approaches, wavelet transformation and Dynamic Time Warping (DTW) techniques have been applied to remove more efficiently the instrument noise and align the signals with respect to the reference. The Euclidean distance of the DTW has been introduced as index to get the best filter parameters configuration. Obtained results, both during the calibration and the investigated real scenario, confirm that at least a GPS signal with a sampling frequency of 1 is needed to track the rider’s behaviour to detect events. In conclusion, with a very cheap hardware setup is possible to monitor riders’ speed and acceleration.