Tata Consultancy Services Limited

Indian Institute of Technology Bombay

Soma Bandyopadhyay

Arpan Pal

Tapas Chakravarty

Arijit Ukil

Rituraj Singh

Rohan Banerjee

Abhijan Bhattacharyya

Sakyajit Bhattacharya

Kayapanda Muthana Mandana

C A Murthy

Ayan Mukherjee

Avik Ghose

Shameemraj Mohinuddin Nadaf

Debayan Mukherjee

Anish Datta

Aniruddha Sinha

Puneet Kumar Agarwal

Sundeep Khandelwal

Munmun Sengupta

Arijit Chowdhury

Arijit Mukherjee

Chirabrata Bhaumik

Tanushyam Chattopadhyay

Snehasis Banerjee

Anirban Dutta Choudhury

Prateep Misra

Chetanya Puri

Ishan Sahu

Tulika Bose

Swarnava Dey

Shreyasi Datta

Vivek Chandel

Himadri Sekhar Paul

Ramesh Kumar Ramakrishnan

Shahnawaz Alam

Dhaval Satish Jani

Satanik Panda

Abhisek Das

Arjit Ukil

Abhijan Bhattacharya

Souvik Maiti

Subhajit Dutta

Trisrota Deb

Riddhi Panse

Krithivasan Ramamritham

Indrajit Bhattachyarya

This disclosure relates generally to method and system for an adaptive filter based learning model for time series sensor signal classification on edge devices. The adaptive filter based learning model for time series sensor signal classification enables automated-computationally lightweight learning (significant reduction in computational resources) and inferring/classification in real-time or near-real-time on CPU/memory/battery life constrained edge devices. The disclosed techniques for time series sensor signal classification on edge devices characterizes the intrinsic signal processing properties of the input time series sensor signals using linear adaptive filtering and derivative spectrum to efficiently construct the adaptive filter based learning model based on standard classification algorithms for time series sensor signal classification.

Adaptive filter based learning model for time series sensor signal classification on edge devices

This disclosure relates generally to a system and a method for mitigating generalization loss in deep neural network for time series classification. In an embodiment, the disclosed method includes compute an entropy of a timeseries training dataset, and a mean and a variance of the entropy and a regularization factor is computed. A plurality of iterations are performed to dynamically adjust the learning rate of the deep Neural Network (DNN) using a Mod-Adam optimization, and obtain a network parameter, and based on the network parameter, the regularization factor is updated to obtain an updated regularized factor. The learning rate is adjusted in the plurality of iterations by repeatedly updating the network parameter based on a variation of a generalization loss during the plurality of iterations. The updated regularized factor of the current iteration is used for adjusting the learning rate in a subsequent iteration of the plurality of iterations.

System and method for mitigating generalization loss in deep neural network for time series classification

This disclosure relates generally to method and system for time series classification. Conventional methods for time-series classification requires substantial amount of annotated data for classification and label generation. The disclosed method and system are capable of generating accurate labels for time-series data by utilizing a small amount of representative data for each class. In an embodiment, the disclosed method generates a time-series data synthetically and associated labels by using a portion of the representative time-series data in each iteration, and self-correcting the generated labels based on a determination of quality of the generated labels using label quality checker models.

System and method for label generation for timeseries classification

This disclosure relates generally to methods and systems for unobtrusive digital health assessment of high risk subjects, wherein bio-markers pertaining to a disease are identified automatically using physical activity and physiology monitoring on a continuous basis. Identification of bio-markers in the medical domain is conventionally dependent on insights derived from medical tests which are obtrusive in nature. Systems and methods of the present disclosure integrate physical characteristics, lifestyle habits and prevailing medical conditions with monitored physical activities and physiological measurements to assess health of high risk subjects. Systems and methods of the present disclosure also enable automatic generation of control class and treatment class that may be effectively used for health assessment.

Systems and methods for unobtrusive digital health assessment

This disclosure relates generally to a unified platform for domain adaptable human behaviour inference. The platform provides a unified, low level inference and high level inference of domain adaptable human behaviour inference. The low level inferences include cross-sectional analysis techniques to infer location, activity, physiology. Further the high inference that provide useful and actionable for longitudinal tracking, prediction and anomaly detection is performed based on several longitudinal analysis techniques that include welch analysis, cross-spectrum analysis, Feature of interest (FOI) identification and time-series clustering, autocorrelation-based distance estimation and exponential smoothing, seasonal and non-seasonal models identification, ARIMA modelling, Hidden Markov models, Long short term memory (LSTM) along with low level inference, human meta-data and application domain knowledge. Further the unified human behaviour inference can be obtained across multiple domains that include health, retail and transportation.

Unified platform for domain adaptable human behaviour inference

This disclosure relates generally to physiological monitoring, and more particularly to feature set optimization for classification of physiological signal. In one embodiment, a method for physiological monitoring includes identifying clean physiological signal training set from an input physiological signal based on a Dynamic Time Warping (DTW) of segments associated with the physiological signal. An optimal features set is extracted from a clean physiological signal training set based on a Maximum Consistency and Maximum Dominance (MCMD) property associated with the optimal feature set that strictly optimizes on the objective function, the conditional likelihood maximization over different selection criteria such that diverse properties of different selection parameters are captured and achieves Pareto-optimality. The input physiological signal is classified into normal signal components and abnormal signal components using the optimal features set.

System and method for physiological monitoring and feature set optimization for classification of physiological signal

Conventional hierarchical time-series clustering is highly time consuming process as time-series are characteristically lengthy. Moreover, finding right similarity measure providing best possible hierarchical cluster is critical to derive accurate inferences from the hierarchical clusters. Method and system for Auto Encoded Compact Sequences (AECS) based hierarchical time-series clustering that enables compact latent representation of time-series using an undercomplete multilayered Seq2Seq LSTM auto encoder followed by generating of HCs using multiple similarity measures is disclosed. Further, provided is a mechanism to select the best HC among the multiple HCs on-the-fly, based on an internal clustering performance measure of Modified Hubert statistic &#x3c4;. Thus, the method provides time efficient and low computational cost approach for hierarchical clustering for both on univariate and multivariate time-series. AECS approach provides a constant length sequence across diverse length series and hence provides a generalized approach.

Method and system for hierarchical time-series clustering with auto encoded compact sequence (AECS)

In many real-life applications, ample amount of examples from one class are present while examples from other classes are rare for training and learning purposes leading to class imbalance problem and misclassification. Methods and systems of the present disclosure facilitate generation of an extended synthetic rare class super dataset that is further pruned to obtain a synthetic rare class dataset by maximizing similarity and diversity in the synthetic rare class dataset while preserving morphological identity with labeled rare class training dataset. Oversampling methods used in the art result in cloning of datasets and do not provide the needed diversity. The methods of the present disclosure can be applied to classification of noisy phonocardiogram (PCG) signals among other applications.

Synthetic rare class generation by preserving morphological identity

A method and system for a feature subset-classifier pair for a classification task. The classification task corresponds to automatically classifying data associated with a subject(s) or object(s) of interest into an appropriate class based on a feature subset selected among a plurality of features extracted from the data and a classifier selected from a set of classifier types. The method proposed includes simultaneously determining the feature subset-classifier pair based on a relax-greedy {feature subset, classifier} approach utilizing sub-greedy search process based on a patience function, wherein the feature subset-classifier pair provides an optimal combination for more accurate classification. The automatic joint selection is time efficient solution, effectively speeding up the classification task.

Method and system for joint selection of a feature subset-classifier pair for a classification task

Conventional systems and methods of classifying heart signals include segmenting them which can fail due to the presence of noise, artifacts and other sounds including third heart sound 'S3', fourth heart sound 'S4', and murmur. Heart sounds are inherently prone to interfering noise (ambient, speech, etc.) and motion artifact, which can overlap time location and frequency spectra of murmur in heart sound. Embodiments of the present disclosure provide parallel implementation of Deep Neural Networks (DNN) for classifying heart sound signals (HSS) wherein spatial (presence of different frequencies component) filters from Spectrogram feature(s) of the HSS are learnt by a first DNN while time-varying component of the signals from MFCC features of the HSS are learnt by a second DNN for classifying the heart sound signal as one of normal sound signal or murmur sound signal.

Parallel implementation of deep neural networks for classifying heart sound signals

Systems and methods for detecting an anomaly in a cardiovascular signal using hierarchical extremas and repetitions. The traditional systems and methods provide for some anomaly detection in the cardiovascular signal but do not consider the discrete nature and strict rising and falling patterns of the cardiovascular signal and frequency in terms of hierarchical maxima points and minima points. Embodiments of the present disclosure provide for detecting the anomaly in the cardiovascular signal using hierarchical extremas and repetitions by smoothening the cardiovascular signal, deriving sets of hierarchical extremas using window detection, identifying signal patterns based upon the sets of hierarchical extremas, identifying repetitions in the signal patterns based upon occurrences and randomness of occurrences of the signal patterns and classifying the cardiovascular signal as anomalous and non-anomalous for detecting the anomaly in the cardiovascular signal.

Systems and methods for detecting anomaly in a cardiovascular signal using hierarchical extremas and repetitions

The present disclosure addresses the technical problem of information loss while representing a physiological signal in the form of symbols and for recognizing patterns inside the signal. Thus making it difficult to retain or extract any relevant information which can be used to detect anomalies in the signal. A system and method for anomaly detection and discovering pattern in a signal using morphology aware symbolic representation has been provided. The system discovers pattern atoms based on the strictly increasing and strictly decreasing characteristics of the time series physiological signal, and generate symbolic representation in terms of these pattern atoms. Additionally the method possess more generalization capability in terms of granularity. This detects discord/abnormal phenomena with consistency.

Method and system for pattern recognition in a signal using morphology aware symbolic representation

Non-invasive methods for accurately classifying Coronary Artery Disease (CAD) is a challenging task. In the present disclosure, a two stage classification is performed. In the first stage of classification, a metadata based rule engine is utilized to classify a subject into one of a confirmed CAD subject, a CAD subject and a non-CAD subject. Here, a set of optimal parameters are selected from a set of metadata associated with the subject based on a difference in frequency of occurrence of the CAD among a disease population and a non-disease population. Further, an optimal threshold associated with each optimal parameter is calculated based on an inflexion based correlation analysis. Further, the CAD subject, classified by the metadata based rule engine is further reclassified in a second stage by utilizing a set of cardiovascular signal into one of the CAD subject and the non-CAD subject.

System and method for classification of coronary artery disease based on metadata and cardiovascular signals

Embodiments herein provide a method for imputing sensor data, in a sensor data sequence with missing data based on the semantics learning, where semantics is defined by the constraints of the sensor data features. A candidate value for imputation is determined based on sensor data of corresponding instances of time instants of the sensor data sequence using learning based on semantics of features of the sensor data sequence with missing data. The nearest neighbors search has been applied in similar response data sequence using the data values corresponding to the time instant of missing data in sensor data sequence. In case similar response data sequence is not available imputation is performed based on the distribution pattern of missing data.

Method for imputing missed data in sensor data sequence with missing data

This disclosure relates generally to the use of distributed system for computation, and more particularly, relates to a method and system for optimizing computation and communication resource while preserving security in the distributed device for computation. In one embodiment, a system and method of utilizing plurality of constrained edge devices for distributed computation is disclosed. The system enables integration of the edge devices like residential gateways and smart phone into a grid of distributed computation. The edged devices with constrained bandwidth, energy, computation capabilities and combination thereof are optimized dynamically based on condition of communication network. The system further enables scheduling and segregation of data, to be analyzed, between the edge devices. The system may further be configured to preserve privacy associated with the data while sharing the data between the plurality of devices during computation.

System and method for distributed computation using heterogeneous computing nodes

Growing community of inventors

Kolkata, India

Soma Bandyopadhyay