Tata Consultancy Services Patente

State of the art methods require size of DL model, or its gradients be less than maximum data item size of storage used as a communication channel for model training with serverless platform. Embodiments of the present disclosure provide a method and system for training large DL models via serverless architecture using communication channel when the gradients are larger than maximum size of one data item allowed by the channel. Gradients that are generated by each worker during current training instance, are chunked into segments and stored in the communication channel. Corresponding segments of each worker are aggregated by aggregators and stored back. Each of the aggregated corresponding segments are read by each worker to generate an aggregated model to be used during successive training instance. Optimization techniques are used for reading-from and writing-to the channel resulting in significant improvement in the performance and the cost of training.

This disclosure relates to time sensitive processing of TCP segments into application layer messages in FPGA. Certain applications such as "stock market" or "ticket booking system" require a time sensitive ordering of the transaction, as the timing of arrival of transaction (packet) will impact the result, wherein the time sensitive ordering occurs when a first packet reaching the application network is processed first or the processing of packets by the server is guaranteed in the order of packets received. However, the existing systems do not honor the time due to the layered network stack. The disclosure is a design and implementation of a middleware framework on FPGA platform which delivers messages to the application in the order in which they arrive. The disclosure enables time sensitive analysis of each message of the TCP segment based on the session-based information to re-assemble the plurality of messages in a time-sensitive queue.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

State of art techniques that utilize spatial association based Table structure Recognition (TSR) have limitation in selecting minimal but most informative word pairs to generate digital table representation. Embodiments herein provide a method and system for TSR from an table image via deep spatial association of words using optimal number of word pairs, analyzed by a single classifier to determine word association. The optimal number of word pairs are identified by utilizing immediate left neighbors and immediate top neighbors approach followed redundant word pair elimination, thus enabling accurate capture of structural feature of even complex table images via minimal word pairs. The reduced number of word pairs in combination with the single classifier trained to determine the word associations into classes comprising as same cell, same row, same column and unrelated, provides TSR pipeline with reduced computational complexity, consuming less resources still generating more accurate digital representation of complex tables.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

The disclosure herein relates to methods and systems for generating an end-to-end de-smoking model for removing smoke present in a video. Conventional data-driven based de-smoking approaches are limited mainly due to lack of suitable training data. Further, the conventional data-driven based de-smoking approaches are not end-to-end for removing the smoke present in the video. The de-smoking model of the present disclosure is trained end-to-end with the use of synthesized smoky video frames that are obtained by source aware smoke synthesis approach. The end-to-end de-smoking model localize and remove the smoke present in the video, using dynamic properties of the smoke. Hence the end-to-end de-smoking model simultaneously identifies the regions affected with the smoke and performs the de-smoking with minimal artifacts. localized smoke removal and color restoration of a real-time video.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Traditional infrastructure monitoring systems have the disadvantage that they either fail to consider dependency between metrices or consider the dependency only at an abstract level, which adversely affects efficiency with which a performance assessment of the infrastructure being monitored can be carried out. Another disadvantage of the existing systems is that manual intervention is required at different stages of the infrastructure monitoring. The disclosure herein generally relates to infrastructure monitoring, and, more particularly, to a method and system for identifying correlation between metrices in the infrastructure. The system uses a hybrid correlation approach which considers determining a dominant correlation from among a direct correlation and a chained correlation that may exist between each pair of metrices being considered at a time for determining the correlation.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

This disclosure relates generally to action recognition and more particularly to system and method for real-time radar-based action recognition. The classical machine learning techniques used for learning and inferring human actions from radar images are compute intensive, and require volumes of training data, making them unsuitable for deployment on network edge. The disclosed system utilizes neuromorphic computing and Spiking Neural Networks (SNN) to learn human actions from radar data captured by radar sensor(s). In an embodiment, the disclosed system includes a SNN model having a data pre-processing layer, Convolutional SNN layers and a Classifier layer. The preprocessing layer receives radar data including doppler frequencies reflected from the target and determines a binarized matrix. The CSNN layers extracts features (spatial and temporal) associated with the target's actions based on the binarized matrix. The classifier layer identifies a type of the action performed by the target based on the features

Zusammenfassung wird geladen …

This disclosure relates generally to method and system for low sampling rate electrical load disaggregation. At low sampling rates, disaggregation of energy load is challenging due to unavailability of events and signatures of the constituent loads. The disclosed energy disaggregation technique receives aggregated load data from a utility meter and sequentially obtains training data for determining disaggregated energy load at low sampling rate. Dictionaries are used to characterize the different loads in terms of power values and time of operation. The obtained dictionary coefficients are treated as graph signals and graph smoothness is used for propagating the coefficients from the training phase to the test phase by formulating an optimization model. The derivation of the optimization model identifies the load of interest and estimate their power consumption based on optimization model constraints. This method achieves accuracy greater than 70% for the loads of interest at low sampling rates.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

This disclosure relates to portable speckle imaging system and method for automated speckle activity map based dynamic speckle analysis. The embodiments of present disclosure herein address unresolved problem of capturing variations in speckle patterns where noise is completely removed and dependency on intensity of variations in speckle patterns is eliminated. The method of the present disclosure provides a correlation methodology for analyzing laser speckle images for applications such as seed viability, fungus detection, surface roughness analysis, and/or the like by capturing temporal variation from frame to frame and ignoring the intensity of speckle data after denoising, thereby providing an effective mechanism to study speckle time series data. The system and method of the present disclosure performs well in terms of time efficiency and visual cues and requires minimal human intervention.

This disclosure relates generally to methods and systems for real time unobtrusive monitoring of physiological signals of a subject confined to a bed. Output of single channel continuous wave (CW) radars are dependent upon distance of the subject from the radar. Dual channel IQ radars are more accurate but are costly and availability in the market is a constraint. The present disclosure provides a cheap and easily replicable pseudo IQ radar based system. The pseudo IQ radar comprises two CW radars placed at a calibrated distance from each other such that optimum points of one CW radar spatially overlaps null points of the other CW radar and phase imbalance is suppressed. Three such pseudo IQ radars are positioned in a predetermined configuration around the subject being monitored. A Supervised Complex Signal Demodulation (SCSD) method configured to suppress amplitude and DC imbalance is also provided for evaluating the physiological signals.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

This disclosure relates generally to method and system for enhancing Quality of experience (QoE) for rendering live interactive videos in real-time. The conventional methods utilize an additional field in packet header to determine which packet sequences were lost, so that the receiver could request a retransmission. Due to unpredictable delay and loss, the receiver may not get the end-of-frame packet which has been sent using best-effort. The disclosed method and system introduced an additional indicator to identify an expected maximum number of packets transmitted for each frame, and further determine the distribution of loss so that the rendering unit can intelligently decide on whether to allow the frame to display, or to drop it without using any additional retransmission or error-concealment strategy, thereby enhancing the QoE of reception at the receiver.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …