Abstract: A system for hallucination detection in text generating systems is provided. The system includes processing circuitry that accesses a query and determines a first query-type associated therewith. The processing circuitry accesses a proposed query response for the query. Based on the proposed query response, the processing circuitry generates a set of parameter values for a set of parameters. Based on the set of parameter values and a first set of adaptive weights associated with the set of parameters, the processing circuitry determines a hallucination magnitude for the proposed query response. Based on the hallucination magnitude and a hallucination threshold associated with the first query-type, the processing circuitry generates a hallucination indicator indicative of a presence or an absence of hallucination in the proposed query response.

Abstract: This disclosure relates to a page management method and system thereof. The method includes categorizing, using a first ML model, a set of pages into at least one of a plurality of strands, based on a set of first parameters. Further, the method includes classifying each of the set of pages, using a second ML model, into one of a plurality of sub-strands, based on an associated set of second parameters and the associated strand from the plurality of strands. Further, the method includes determining, for each of the set of pages, a score, based on the associated sub-strand, the weight assigned to each of the subset of second parameters of the associated sub-strand, and values of each of the subset of second parameters of the associated sub-strand. Further, the method includes performing an action on at least one of the set of pages based on the determined score.

Abstract: A system for three-dimensional (3D) reconstruction of interiors of a closed container (e.g., a refrigerator) is provided. A track is mounted inside the refrigerator. A robot is configured to navigate along the track. The robot may include control modules and puller and pusher modules that are magnetically coupled to the control modules. The puller module may pull the control modules along the track, whereas the pusher module may push the control modules along the track. Each control module may include an image sensor that captures images of the physical space inside the refrigerator. The track is mounted such that by navigating along the track, the robot (e.g., the image sensors) can capture an entire scene of a physical space inside the refrigerator from different angles and locations within the refrigerator. Based on the captured images, a 3D representation of the physical space is generated.

Abstract: Adaptive person re-identification technique is provided in the present disclosure. A reference feature profile is created for each individual present in a multi-camera facility. The reference feature profile is indicative of a reference appearance (e.g., apparel and/or accessories) of the individual. In a real-time video frame, a user and a user appearance are detected. Further, a current feature profile is generated for the user based on the detected appearance. The reference feature profile of the detected user is then compared with the current feature profile to detect an appearance-change event. The appearance-change event may be addition/removal of an apparel, addition/removal of an accessory, or a combination thereof. If the appearance-change event is validated, the reference feature profile of the user is updated with the current feature profile. The updated reference feature profile is then utilized for re-identification.

Abstract: A method for access management in graph-based model is provided. The graph-based model comprises a plurality of nodes and an overlay structure comprising processing logic. The overlay structure is associated with one or more nodes of the plurality of nodes. Processing circuitry determines a first node group of the graph-based model. The first node group comprises at least one node. The processing circuitry associates a first contract with the first node group such that the first contract is configured to act as a proxy for one or more nodes within the first node group in relation to requests from outside the first node group. The processing circuitry receives a stimulus and a context associated therewith. The stimulus is associated with the first contract. The processing circuitry maps the stimulus to the first contract to determine an access response. The processing circuitry processes the stimulus based on the access response.

Abstract: This disclosure relates to method and system for facilitating legacy code transformation. The method includes receiving legacy code data and natural language document from one or more data sources. Each of the one or more data sources is one of an external data source or an internal data source. Further, the method includes generating a first natural language output based on the legacy code data through a first LLM, and a second natural language output based on the natural language document through a second LLM. Further, the method includes fine-tuning one of the first LLM or the second LLM based on the first natural language output and the second natural language output, through a third LLM. Further, the method includes generating a natural language specification document corresponding to the legacy code data based on the first natural language output and the second natural language output through the third LLM.

Abstract: A method for access management in graph-based model is provided. The graph-based model comprises a plurality of nodes and an overlay structure comprising processing logic. The overlay structure is associated with one or more nodes of the plurality of nodes. Processing circuitry determines a first node group of the graph-based model. The first node group comprises at least one node. The processing circuitry associates a first contract with the first node group such that the first contract is configured to act as a proxy for one or more nodes within the first node group in relation to requests from outside the first node group. The processing circuitry receives a stimulus and a context associated therewith. The stimulus is associated with the first contract. The processing circuitry maps the stimulus to the first contract to determine an access response. The processing circuitry processes the stimulus based on the access response.

Abstract: An overlay system is provided that includes a storage element and processing circuitry coupled thereto. The storage element stores an executable graph-based model having a plurality of run-time bi-directional nodes and a plurality of run-time connection links. Each run-time bi-directional node includes a node template and a node instance. The processing circuitry receives a stimulus and identifies a first run-time bi-directional node based on the stimulus. The processing circuitry determines a first run-time connection link based on the identification of the first run-time bi-directional node. The first run-time bi-directional node is coupled to a second run-time bi-directional node via the first run-time connection link. The processing circuitry further identifies the second run-time bi-directional node and executes the operation associated with the stimulus based on the first run-time bi-directional node, the second run-time bi-directional node, and the first run-time connection link.

Abstract: An overlay system is provided that includes a storage element and processing circuitry coupled thereto. The storage element stores an executable graph-based model having various active nodes. The processing circuitry may receive a stimulus associated with the overlay system and identify, based on the stimulus, an active node. Further, the processing circuitry may generate a proxy node for the active node. The proxy node includes a reference to the active node. The processing circuitry may implement the proxy node in the executable graph-based model by way of an actor interface. Further, the processing circuitry may store the first proxy node in a primary storage of the storage element. Additionally, as a response to the first stimulus, the processing circuitry may unload the first active node from the executable graph-based model.

Abstract: A system for time-series message management using directed property graphs is provided. A directed property graph is generated using various messages. The directed property graph comprises various message nodes and attribute nodes, with each message node representing a message, having some attributes of the message associated as node properties thereof, and being associated with attribute nodes that represent shared attributes of the message. To execute a query indicative of time-series computation functions, one or more attribute nodes are identified in the directed property graph. Based on the identified attribute nodes, a set of message nodes is determined. Further, from node properties of each determined message node, one or more message timestamps are identified. Based on the identified message timestamps, the time-series computation functions are executed.

Abstract: This disclosure relates to method and system for facilitating legacy code transformation. The method includes receiving legacy code data and natural language document from one or more data sources. Each of the one or more data sources is one of an external data source or an internal data source. Further, the method includes generating a first natural language output based on the legacy code data through a first LLM, and a second natural language output based on the natural language document through a second LLM. Further, the method includes fine-tuning one of the first LLM or the second LLM based on the first natural language output and the second natural language output, through a third LLM. Further, the method includes generating a natural language specification document corresponding to the legacy code data based on the first natural language output and the second natural language output through the third LLM.

Abstract: The present disclosure facilitates optimized storage of historical data. A graph is generated using various messages. For each message, a message node is instantiated in the graph. In a message-driven architecture, processing of a message leads to the generation of a data node in the graph. The data node represents an output associated with the message processing. In response to the data node generation, another message node is instantiated in the graph. The other message node represents a message that is indicative of the generation of the data node and comprises data and transactional information associated with the data node. Further, a history node is created linking the data node and the two message nodes to indicate that the data node is an output associated with the message processing and the newly generated message node stores a change associated with the data node.

Abstract: A system for message management using directed property graphs is provided. A directed property graph is generated using various messages. To generate the graph, for each message, a message node representing the message is instantiated. Each message has various attributes. Some message attributes are associated as node properties of the message node. Further, one or more shared attributes are derived for each message and corresponding attribute nodes are instantiated. One or more attributes are associated as node properties of these attribute nodes. Further, each message node is coupled to a corresponding attribute node by way of an edge. Each edge has edge attributes associated as edge properties. The edge attributes may be derived from node properties of the message node and the corresponding attribute node and may indicate an association therebetween. The generated graph facilitates query response generation.

Abstract: The disclosure relates to system and method for continuously tracking humans in an area. The method includes receiving video data of the area from overhead cameras. Each of overhead cameras includes Field of View (FoV), FoV includes overlapping region and non-overlapping region, and overlapping region corresponds to region of intersection between at least two FoVs. The method further includes detecting presence humans in first FoV through object detection and classification models; for each human of humans, assigning unique global identity (ID) corresponding to human in first FoV, and reassigning unique global ID to human when human moves from first FoV to second FoV through overlapping region between first FoV and second FoV using weighted combination of resource assignment algorithm, intersection-over-union (IOU) based track detection, and velocity and direction estimation of subsequent frame of video data; and continuously tracking, in real-time, each of humans in the area through unique global ID.

Abstract: A system and/or a method for a cognitive platform for autonomous data orchestration, comprising edge computing devices, data platforms and cognitive computing engine. The cognitive computing engine is configured to receive digital events either generated by external computing devices or within the cognitive computing engine. Context and state information is accessed from a central intelligence store to determine the intent of the received digital events. A target entity is identified based on at least one of the determined intent. A data package is composed for the identified at least one target entity, transformed to a natural language text and communicated to the identified at least one target entity, inciting the target entity to take action on the package delivered. The central intelligence store is updated new context and intents for subsequent digital events making it an autonomous learning platform.

Abstract: This disclosure relates to a method and system for sanitization of sensitive data. The method includes analyzing sensitive data present within a page of an application based on a deterministic algorithm. Further, the method includes classifying the sensitive data into a high risk sensitive data and a low risk sensitive data based on a ML classification algorithm. For the one or more sensitive data classified as the high risk sensitive data, the method further includes performing a destructive sanitization on each of the high risk sensitive data. For the one or more sensitive data classified as the high risk sensitive data, the method further includes performing a non-destructive sanitization on each of the high risk sensitive data.

Abstract: An overlay system is provided that includes processing circuitry and a storage element that stores an executable graph-based model including various active nodes, various history message nodes, and various history overlay nodes. Each active node is associated with one or more history overlay nodes that facilitate the creation and maintenance of one or more history message nodes associated therewith. The processing circuitry receives a contextualized stimulus and identifies an active node and one or more history message nodes in the executable graph-based model based on the context. The processing circuitry creates one or more history nodes (e.g., one or more historical versions of the corresponding active node) based on the one or more history message nodes. Further, the processing circuitry executes an operation associated with the stimulus based on the identified active node, the one or more history overlay nodes, and the created one or more history nodes.

Abstract: The disclosure relates to system and method for smart test execution in a network test automation platform. The method includes retrieving set of parameter values for each of plurality of test cases in test suite corresponding to set of risk parameters based on latest release of product code. The method further includes calculating risk index for each of plurality of test cases based on set of parameter values and predefined set of weightage scores. The method further includes determining execution order of plurality of test cases based on calculated risk index. The method further includes executing set of priority test cases from plurality of test cases based on execution order and predefined threshold risk escape value. The method further includes dynamically rendering in real-time, one or more live charts based on current risk escape value upon execution of the set of priority test cases via GUI.

Abstract: A system for improving performance of variational quantum classifiers (VQCs) using large language models (LLMs) is provided. Features and classes are extracted from a classification dataset. Weights are assigned to the features based on contribution of the features in class identification. Further, segregation parameters are determined for the features such that a segregation parameter indicates a degree of separation between two classes based on a correlation between two features. A prompt is generated based on the weights, the segregation parameters, current ansatz and feature map of a VQC, a performance metric of the VQC for the current ansatz and feature map, details of hardware executing the VQC, and a number of qubits of the VQC. Using an LLM, a response to the prompt is generated. Further, using the response, the ansatz and the feature map are updated. The update continues until the performance metric is within a desired range.

Abstract: This disclosure relates to method and system for framework agnostic smart test orchestration in network test automation in a network. The method includes receiving input data for executing one or more test scripts to test one or more functions on one or more devices in the network. The input is received from a user device via a web-based application. The method comprises parsing the input data and selecting an execution module from a plurality of execution modules for executing the one or more test scripts on the one or more devices based on the parsed input data. The method comprises executing the one or more test scripts to test the one or more functions on the one or more devices using the selected execution module.