Tata Consultancy Services Patente

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

The disclosure generally relates to methods and systems for generating synthetic microstructure images of a material with desired features. Conventional techniques that make use of unsupervised deep generative models has no control on the generated microstructure images with specific, desired set of features. The present disclosure generates the synthetic microstructure images of the material with desired feature, by using a variational autoencoder defined with a style loss function. In the first step, the variational autoencoder is trained to learn latent representation of microstructure image of the material. In the second step, some of the dimensions of learned latent representation is interpreted as physically significant features. In the third and last step, the latent representation required for getting the desired features is appropriately sampled based on the interpretation to generate the synthetic microstructure images of the material with desired features.

Zusammenfassung wird geladen …

Currently available live media distribution systems have to continuously maintain multiple P2P connections which further increases the computational load of robot. Present disclosure provides a method and a system for managing multimedia exchange in tele-robotics using dew computing. The system first identifies a dew signaling server wherein a tele-presence robot device present in enterprise network offer to act as dew signaling server. The system then establishes first peer-to-peer (P2P) connection between dew interface and dew signaling server. Thereafter, system establishes second P2P connection between dew signaling server and cloud media manager present in public cloud server. Further, system identifies computing device present in enterprise network as dew media manager. The system then establishes third P2P connection between dew media manager and the cloud media manager via the cloud signaling server. The dew media manager then perform multimedia exchange within one or more enterprise user devices connected to enterprise network.

This disclosure relates generally to method and system for determining label value for labels in unstructured documents. Typical systems have challenge in understanding variations in layout of unstructured documents and extract information therefrom. The disclosed method and system facilitate systematically identifying sections and bounding boxes in the page images, taking image portion of the bounding boxes and extracting labels and label values therefrom. In case the label values are not present in the same bounding box having the label, the neighboring labels are examined for the matching label values. The system also obtains label-label value pairs from the document by utilizing a trained deep learning model, and compares the output with the label-label value pairs extracted earlier. An aggregated confidence score is assigned to the text in the bounding box.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Conventionally, most equipment manipulating object have been huge with rotatable pivot platforms, multiple roller conveyors, telescopic boom conveyors, hydraulic cranes, etc. These systems may have higher lifecycle cost due to their size and bill of materials (BoM) count and have limited themselves to handling either loading or unloading and lack to perform both the operations. Present disclosure provides apparatus for organized loading and unloading of objects. The apparatus addresses the issue of single equipment for both loading, unloading, and further automatically organizes the objects in bulk thus ensuring safety, without damages. The objects are placed optimally while being unloaded onto pallets or other carriers. While loading into a container, the objects are loaded to a greater fill volume. These operations of multiple objects handling in one go are enabled through various belt conveyors, manipulators, vacuum generators, suction panels, object gripping and conveying platforms, and object gripping systems.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

This disclosure relates to system and method to reconstruct human motion for mobile robot teleoperation using shared control. The method of the present disclosure acquire an input feed of a human operator to perform a task with assistance in a remote environment using shared control. The mobile robot reconstructs to follow a trajectory of the human operator towards the intended goal in the remote environment. The mobile robot determines at least one goal intended by the human operator based on a previously occurred state, a current kinematic state and a future trajectory of the human operator and a known position of the plurality of goals. The model predictive control generates at least one instruction to control the movement of the mobile robot to perform at least one of following the trajectory of the human operator and reaching the operator intended goal based on a joint angle position and a velocity.

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 …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Retailers need information about their competitor's pricing and promotions and marketing collaterals are one of the most vital sources of this information. Conventional approaches for extracting product names from marketing collaterals depends on large volume of data repositories and complex machine learning based approaches. The present disclosure extracts product name blocks from marketing collaterals using image processing techniques. The inputs to the present disclosure are seed words and the marketing collateral. A plurality of word level text regions from the image and text value are obtained. Further, a plurality of text characteristics corresponding to each of the plurality of word level text regions are extracted and matching seed word regions are obtained. Further a plurality of meaningful text blocks and a plurality of seed blocks are obtained. Finally, a plurality of product names are extracted using a matrix based product name detection technique and updated in the product dictionary.

Existing electricity consumption prediction approaches depend largely on data-based models, which may be statistical techniques or more precisely time-series predictions that predict based on the auto-regressive nature of the load curve with a few external variables at best, such as calendar events and ambient temperature. While these models are effective to a certain degree for the overall grid level requirements, they may not be able to predict disruptive changes that may happen over longer periods of time, such as the demographic shifts, etc. Method and system disclosed herein predict the electricity consumption by taking into consideration various parameters associated with such disruptive changes, and then predict the electricity consumption in a target area by aggregating the electricity consumption predicted at the agent level.

Zusammenfassung wird geladen …

Current technologies analyze electrocardiogram (ECG) signals for a long duration, which is not always a practical scenario. Moreover the current scenarios perform a binary classification between normal and Atrial Fibrillation (AF) only, whereas there are many abnormal rhythms apart from AF. Conventional systems/methods have their own limitations and may tend to misclassify ECG signals, thereby resulting in an unbalanced multi-label classification problem. Embodiments of the present disclosure provide systems and methods that are robust and more efficient for classifying rhythms for example, normal, AF, other abnormal rhythms and noisy ECG recordings by implementing a spectrogram based noise removal that obtains clean ECG signal from an acquired single-lead ECG signal, an optimum feature selection at each layer of classification that selects optimum features from a pool of extracted features, and a multi-layer cascaded binary classifier that identifies rhythms in the clean ECG signal at each layer of the classifier.

The present disclosure discloses device and method for detecting objects placed at multiple ranges from vehicle. Images of the objects may be captured by an image capturing unit housed in the vehicle. The image may be splitted into plurality of sub-images. Further, one or more features may be extracted from the plurality of sub-images. Further, each of the plurality of sub-images may be simultaneously processed for computing gradients associated with the plurality of sub-images. Further, a cell histogram may be created by casting weighted vote for an orientation based histogram channel based on values associated with the gradient. The gradients computed may be normalized by grouping the cells in spatial blocks. Further, a Support vector Machine (SVM) linear classifier may be applied on the plurality of sub-images in order to classify the near object and the far object in a category of a pedestrian or a vehicle.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Rotary knifes/cutters play an important role in manufacturing of finished products. The rotary cutters tend to lose their cutting material over time. Hence to compensate, pressure applied by cylinder over rotary cutter needs to be changed. But this change in pressure needs to be optimum as too high pressure can lead to loss of material and too low pressure can stop cutting operation. Present application provides methods and systems for real time estimation of pressure change requirements for rotary cutters. The system first determines minimum and maximum usage limit for rotary cutter based on historical rotary cutter usage data and real-time pressure value using first trained model. The system, upon determining that minimum usage limit is reached, determines time for next pressure change based on physical parameters using second trained model. Thereafter, system compares estimated time with estimated maximum usage limit and displays notification to change pressure based on comparison.

Reinforcement Learning agent interacting with a real-world building to determine optimal policy may not be viable due to comfort constraints. Embodiments of the present disclosure provide multi-deep agent RL for dynamically controlling electrical equipment in buildings, wherein a simulation model is generated using design specification of (i) controllable electrical equipment (or subsystem) and (ii) building. Each RL agent is trained using simulation model and deployed in the subsystem. Reward function for each subsystem includes some portion of reward from other subsystem(s). Based on reward function of each RL agent, each RL agent learns an optimal control parameter during execution of RL agent in subsystem. Further, a global optimal control parameter list is generated using the optimal control parameter. The control parameters in the global optimal control parameters list are fine-tuned to improve subsystem's performance. Information on fine-tuning parameters of the subsystem and reward function are used for training RL agents.

Zusammenfassung wird geladen …

Retail stores have limited visibility of on shelf inventory. Conventional approaches for targeted replenishment are reactive in nature and are also infrastructure and labor heavy. Present disclosure provides systems and methods for identification and replenishment of targeted items on shelves of stores wherein input data pertaining to sales of items is pre-processed and stock keeping unit (SKU) wise optimal bucket size is determined for predicting sales events for individual SKU based on historical events. Top-up requests are generated for each SKU for the planning bucket sizes and further a pick-up list using smart batching of the top-up requests is created based on SKU priorities. The pick-up list and top-up requests are executed to ensure items are topped up at the right time. Further, rate of sales or forecast the rate of sales are continually monitored throughout the day to ensure items are identified for targeted replenishment in retail stores.