A computer-implemented method of enhancing a computer to estimate an uncertainty of an onset of a signal of interest in time-series noisy data. A first mathematical model of first time series data that contains only noise is calculated. A second mathematical model of second time series data that contains the noise and an onset of a signal of interest in the second time series data is calculated. A difference is evaluated between a first combination, being the first mathematical model and the second mathematical model, and a second combination, being the first time series data and the second time series data, wherein evaluating is performed using a generalized entropy metric. A specific time when an onset of the signal of interest occurs is estimated from the difference. An a posteriori distribution is derived for an uncertainty of the specific time at which the onset occurs.

Methods, apparatus, systems, and articles of manufacture are disclosed to improve media identification. An example apparatus includes a hash handler to generate a first set of reference matches by performing hash functions on a subset of media data associated with media to generate hashed media data based on a first bucket size, a candidate determiner to identify a second set of reference matches that include ones of the first set, the second set including ones having first quantities of hits that did not satisfy a threshold, determine second quantities of hits for ones of the second set by matching ones to the hash tables based on a second bucket size, and identify one or more candidate matches based on at least one of (1) ones of the first set or (2) ones of the second set, and a report generator to generate a report including a media identification.

Methods and systems for predicting failure in a cyber-physical system include determining a prediction index based on a comparison of input time series, from respective sensors in a cyber-physical system, to failure precursors. A failure precursor is detected in the input time series, responsive to a comparison of the prediction index to a threshold. A subset of the sensors associated with the failure precursor is determined, based on a gradient of the prediction index. A corrective action is performed responsive to the determined subset of sensors.

An aspect of the present invention more reliably prevents a false detection of lifting of the electronic device, in an information processing device mounted on an electronic device. A first lifting determination section (64A) determines that a mobile terminal (1) has been lifted, in a case where the following conditions (i) and (ii) are satisfied after a manner of change in acceleration detected by an acceleration sensor (11) over time has satisfied a predetermined acceleration condition: (i) a standstill determination section (63) determines that the mobile terminal (1) is in a standstill state; and (ii) a result of detection by a proximity sensor (14) indicates a transition from proximity to non-proximity within a predetermined time range, the predetermined time range being set with reference to a standstill determination completion time point at which the standstill determination section (63) has completed determination.

Systems and methods are disclosed for device movement-based authentication. One method comprises receiving contextual data from one or more sensors of a user device and determining a device movement pattern based on the received contextual data. The determined device movement pattern is compared with a device movement-based signature associated with a user of the user device. If the determined device movement pattern matches the device-movement based signature within a predetermined threshold, the user is authenticated for an electronic transaction. If the determined device movement pattern does not match the device-movement based signature, a notification indicating authentication failure is sent to the user device.

Embodiments of the present invention provide a method and system for timing related tasks in IoT systems, for example, in relation to synchronisation of clocks and timestamping. It is desirable that the method and system is able to withstand external tampering in a manner which does not jeopardise the accuracy and integrity of time related tasks in the IoT systems.

A computer-implemented method, computerized apparatus and computer program product for activity recognition using adaptive window size segmentation of sensor data stream. A data stream generated by one or more sensors is obtained. A frequency analysis of the data in a first segment of the data stream is performed. A size of a second segment is determined based on the frequency analysis. Activity recognition is performed for the second segment by extracting one or more features of the data therein and applying a machine learning process on the extracted features to obtain a classification of the data into an activity class.

An information processing device includes a motion information acquiring section that acquires a plurality of pieces of motion information showing a series of motions of a same kind performed by a plurality of users and a reference timing extracting section that extracts a reference timing based on a plurality of pieces of the motion information. Also included are an image information acquiring section that acquires a plurality of series of image information corresponding to a plurality of pieces of the motion information and a synchronization processing section that synchronizes a plurality of pieces of the image information based on the reference timing.

According to an aspect, there is provided an apparatus (52) for determining whether a subject has entered or exited a building, wherein the apparatus (52) is configured to obtain air pressure measurements from an air pressure sensor (58; 60) associated with the subject, wherein the air pressure measurements comprise a first plurality of air pressure measurement samples; determine one or more of: a first autocorrelation signal for the first plurality of air pressure measurement samples at a first time lag; a distribution of air pressures for the first plurality of air pressure measurement samples; a distribution of air pressure differences for the first plurality of air pressure measurement samples, wherein each air pressure difference is the difference between the value of a first air pressure measurement sample in the first plurality of air pressure measurement samples and the value of an air pressure measurement sample in the first plurality of air pressure measurement samples that is a second time lag before the first air pressure measurement sample; and analyse the determined first autocorrelation signal, the determined distribution of air pressures and/or determined distribution to determine one or more indications of whether the subject has entered or exited a building; and determine whether the subject has entered or exited a building based on the determined one or more indications. A corresponding computer-implemented method and computer program product are also provided.

An abnormality diagnosis apparatus includes a time series data obtaining unit for obtaining test data and variable data, the test data being obtained from the test object over a predetermined period of time and being time series data of a predetermined attribute value that dictates whether an abnormality is present in the test object, and the variable data corresponding to the test data for the predetermined period of time and being time series data regarding a variable affecting the attribute value, a superimposed image generating unit for superimposing a waveform of the test data and a waveform of the variable data to generate superimposed image data, and a determining unit for determining whether an abnormality is present in the test object based on the superimposed image data generated by the superimposed image generating unit.