Abstract: Systems and methods described herein provide techniques for determining sleep state data by processing video data of a subject. Systems and methods may determine a plurality of features from the video data, and may determine sleep state data for the subject using the plurality of features. In some embodiments, the sleep state data may be based on frequency domain features and/or time domain features corresponding to the plurality of features.

Abstract: Methods and systems for model based measurement of a structural parameter characterizing a physical characteristic of a measurement target using a physics based measurement model with a trained Machine Learning (ML) based decorrelation model are presented herein. The physics based measurement model estimates values of multiple parameters characterizing multiple, different structural characteristics of a measurement target. The ML based decorrelation model estimates the value of one parameter of interest characterizing the target under measurement based on the values of the parameters estimated by the physics based measurement model, and optionally, the underlying measurement signals. The ML based decorrelation model is trained to leverage process-based correlation among the parameters estimated by the physics based measurement model to accurately estimate the value of one parameter of interest.

Abstract: Techniques for improved machine learning are provided. An approval request specifying a plurality of values for a plurality of request fields is accessed. A model output is generated based on processing the plurality of values using a machine learning model, wherein the output comprises an approval prediction for the approval request, and a denial reason for the approval request. The model output is outputted, and an updated value for at least one of the plurality of values is received. A second model output is generated based on processing the updated value using the machine learning model, and the second model output is outputted.

Abstract: Techniques for extracting data from conversations across different types of communication channels are disclosed. A system applies a set of rules to extract data from conversations based, at least in part, on a type of communication channel used for conducting the conversation. The system applies a machine learning model to recognize semantic content in conversations. The system divides conversations into conversation segments and classifies the conversation segments based on the semantic content. The system selects conversation segments to be extracted based on the semantic content and the type of communication channel over which a conversation is conducted. The system maps conversation segments from different conversations conducted on different types of communication channels to a same set of transactions.

Abstract: Systems and methods described herein provide techniques for analyzing gait and posture of a subject with respect to control data. The systems and methods, in some embodiments, processes video data, identifies keypoints representing body parts, determines metrics data at a stride level, and compares the metrics data to control data.

Abstract: Systems and methods described herein provide techniques for detecting subject behavior by processing video data using one or more trained models configured to detect subject behavior. The described system processes sets of frames from the video data using different trained models. The system further processes different orientations of the sets of frames. The various outputs from the different trained models and from processing the different orientations of the sets of frames may be combined to then make a final determination as to whether the subject is exhibiting a particular behavior during a particular frame.

Abstract: A clustering model is used to partition a set of historical opportunity profiles representing historical opportunities into a plurality of opportunity profile clusters. A target opportunity profile that includes a plurality of target-opportunity parameters is obtained, and an opportunity profile cluster is identified that meets a matching criterion with the target opportunity profile. The plurality of historical opportunity profiles of the identified opportunity profile cluster are partitioned based on an outcome indicator associated respectively with each of the plurality of historical opportunity profiles, providing a first subset of historical opportunity profiles and a second subset of historical opportunity profiles. An insight generation model is applied to the target opportunity profile along with the first subset and/or the second subset to generate one or more insights for the target opportunity profile.

Abstract: Computer-implemented methods for training a neural network, as well as for implementing audio encoders and decoders via trained neural networks, are provided. The neural network may receive an input audio signal, generate an encoded audio signal and decode the encoded audio signal. A loss function generating module may receive the decoded audio signal and a ground truth audio signal, and may generate a loss function value corresponding to the decoded audio signal. Generating the loss function value may involve applying a psychoacoustic model. The neural network may be trained based on the loss function value. The training may involve updating at least one weight of the neural network.

Abstract: Techniques for improved machine learning are provided. Resident data describing a first condition of a first resident of a residential care facility is received. An optimized treatment plan is selected for the first resident by extracting a first plurality of resident attributes, from the resident data, generating a first approach to remediate the first condition, and generating a first predicted recovery score by inputting the first approach and the first plurality of resident attributes into a trained machine learning model. The optimized treatment plan is implemented for the first resident.

Abstract: Example systems, methods, and computer-program products are described for using multiple AI agents to simulate performer-evaluator interaction and generate constructive feedback to automate pipeline progression. An example method includes a performer agent accessing performer-agent-accessible data potentially relevant to a project. The example method also includes the performer agent generating a performer recommendation to explain, from a performer perspective and with reference to at least some of the performer-agent-accessible data, why the project should advance in status according at least to stored objectives. The example method also includes an evaluator agent generating an evaluator recommendation to explain, from an evaluator perspective and with reference to at least some of the performer-agent-accessible data, why the project should not advance in status according at least to the stored objectives.

Abstract: A storage device generates and uses a swift access table for translating a logical block address (LBA) to a physical address/jumbo block address (JBA) on a memory device. When updating a L2P table, a controller determines that data being updated belongs to a section in an mset that includes sequential data and sets continuity metadata for the section. During read-command processing, the controller determines that the continuity metadata for an associated mset is set to a predefined value and, when an entry for the mset is not loaded in the swift access table, the controller loads the entry in the swift access table. The controller calculates the JBA using a relative mset index and a relative JBA offset with respect to an anchor entry in the swift access table and transfers the JBA to the memory device to read the data.

Abstract: Apparatuses and associated techniques for fall detection and prevention via wearable devices are provided. In one aspect, a wearable device is provided including one or more pressure sensors arranged on a bottom portion of the wearable device to collect pressure data from a bottom of a foot of a user, a wireless transmitter communicatively coupled with the one or more pressure sensors, and a power source coupled with the wireless transmitter.

Abstract: An apparatus, method and computer program product for query optimization in a Relational Database Management System (RDBMS), wherein an optimizer accesses a query expression repository (QER), so that the optimizer learns from previous versions of the queries to improve current and subsequent versions of the queries. The QER stores planning and execution information for QEs from the previous versions of the queries, wherein the QEs comprise table relations, intermediate results and/or final results of operations in the previous versions of the queries. The optimizer searches the QER for QEs from the query execution plans, and uses information from the QEs stored in the QER when optimizing the current and subsequent versions of the queries. The optimizer may also reuses results from the QEs stored in the QER.

Abstract: Provided are systems and methods that leverage machine learning to perform audio editing with improved precision and flexibility. Some example systems utilize a vector-based audio editing representation to condition and control a machine-learned audio editing model. This approach allows for detailed and precise control over audio edits by encoding the edits numerically and processing the encoding with a trained model. The system can handle a variety of edits, including audio generation, removal, transformation, time-shifting, and/or enhancement, thereby providing a comprehensive tool for audio manipulation.

Abstract: Implementing a circuit design for a target integrated circuit includes receiving guide information including placement information for a circuit design from a prior implementation flow. A light incremental flow is performed on the circuit design using the guide information. The light incremental flow generates a placed and routed circuit design that re-uses at least a portion of the placement information from the prior implementation flow.

Abstract: Systems and methods for distributing policy enforcement in a network are disclosed. Embodiments may allow the distribution of enforcement of policies from firewalls to other network elements to allow the offloading of the enforcement of actions of those policies for flows to network elements, where the applicability of those policies to those flows was determined by the firewall. Moreover, embodiments may be adapted to update flows offloaded at the network elements to which they have been offloaded with updated policy actions for changed policy actions associated with those flows.

Abstract: Systems and methods described herein provide techniques for detecting subject behavior by processing video data using one or more trained models configured to detect subject behavior. The described system processes sets of frames from the video data using different trained models. The system further processes different orientations of the sets of frames. The various outputs from the different trained models and from processing the different orientations of the sets of frames may be combined to then make a final determination as to whether the subject is exhibiting a particular behavior during a particular frame.

Abstract: This disclosure relates generally to identification and analysis of regime shift. The identification and analysis of the regime shift includes regime shift identification (RSI), root cause analysis of the identified regime shift and a recommendation unit to rectify the identified regime shift. The disclosure proposes to monitor a system continuously to identify a regime shift at real-time as presence of regime shifts in any system decreases quality of process and products and makes the system less efficient. The regime shift is identified at real-time based on key performance indicators (KPIs), a set of relevant features and real time input data using machine learning techniques. Further the disclosure also proposes techniques for detecting at least one root cause for the identified regime shift and also recommends a rectification action to rectify the identified regime shift based on optimization techniques.

Abstract: A clustering model is used to partition a set of historical opportunity profiles representing historical opportunities into a plurality of opportunity profile clusters. A target opportunity profile that includes a plurality of target-opportunity parameters is obtained, and an opportunity profile cluster is identified that meets a matching criterion with the target opportunity profile. The plurality of historical opportunity profiles of the identified opportunity profile cluster are partitioned based on an outcome indicator associated respectively with each of the plurality of historical opportunity profiles, providing a first subset of historical opportunity profiles and a second subset of historical opportunity profiles. An insight generation model is applied to the target opportunity profile along with the first subset and/or the second subset to generate one or more insights for the target opportunity profile.

Abstract: Methods and systems are provided for reducing contactor freezing. In one example, a system may include a contactor comprising a set of conductors and a phase change material. The phase change material is coupled to a first conductor of the set of conductors and is not positioned between the set of conductors.