Abstract: Systems and techniques for adjusting experiment parameters are illustrated. One embodiment includes a method that defines a joint distribution, wherein the joint distribution corresponds to a combination of a probabilistic model and a point prediction model, and wherein the point prediction model is configured to obtain a measurement of regression accuracy. The method derives an energy function for the joint distribution. The method obtains, from the energy function for the joint distribution, an approximation for a conditional distribution, wherein an output of the point prediction model is a parameter of the approximation. The method determines, from a loss function, at least one training parameter. The method trains the probabilistic based on the at least one parameter to operate as a conditional generative model, wherein the trained probabilistic model follows the conditional distribution. The method applies the trained probabilistic model to a dataset corresponding to a randomized trial.

Abstract: Systems and techniques for adjusting experiment parameters are illustrated. One embodiment includes a method that defines a joint distribution, wherein the joint distribution corresponds to a combination of a probabilistic model and a point prediction model, and wherein the point prediction model is configured to obtain a measurement of regression accuracy. The method derives an energy function for the joint distribution. The method obtains, from the energy function for the joint distribution, an approximation for a conditional distribution, wherein an output of the point prediction model is a parameter of the approximation. The method determines, from a loss function, at least one training parameter. The method trains the probabilistic based on the at least one parameter to operate as a conditional generative model, wherein the trained probabilistic model follows the conditional distribution. The method applies the trained probabilistic model to a dataset corresponding to a randomized trial.

Abstract: The present disclosure relates to systems and methods for determining an engagement profile of a participant by associating electronic activities to a profile. It may generate the engagement profile based on analysis of the electronic activity level. An example implementation may contain the following steps. The system may access for a first record object a plurality of electronic activities linked with the first record object. The system may identify for a participant from the plurality of electronic activities a set of electronic activities including the participant. The system may determine an engagement profile of the participant based on a first number of electronic activities of the set of electronic activities sent by the participant, a second number of the set of electronic activities received by the participant and a temporal distribution of the set of electronic activities. The system may store the engagement profile in one or more data structures.

Abstract: There is provided a method for predicting characteristic information of a target to be recognized. The method comprises: acquiring a plurality of first face images for learning and characteristic information on each first face image; generating a plurality of second face images for learning obtained by synthesizing a mask image with the plurality of first face images for learning by a predetermined algorithm; and training a first neural network by using the plurality of second face images for learning as input data for learning and characteristic information as label data for each second face image corresponding to one of the first face images.

Abstract: A program is executed in a first mode of operation in a controlled environment in accordance with normal operations without malicious behavior. An acceptable behavior model is generated based on a plurality of sequences of events that occur during the normal operation of the program. The acceptable behavior model is indicative of normal behavior of the program that occurs during the normal operation. Then the program is executed in a second mode of operation in an operational environment. An operational sequence of events (determined during the second mode of operation) is compared with the acceptable behavior model. When there is a match between the operational sequence of events and the acceptable behavior model, execution in the second mode of operation continues. When there is not a match between the operational sequence of events and the acceptable behavior model, execution in the second mode of operation is halted.

Abstract: Embodiments presented herein describe techniques for generating a linguistic model of input data obtained from a data source (e.g., a video camera). According to one embodiment of the present disclosure, a sequence of symbols is generated based on an ordered stream of normalized vectors generated from the input data. A dictionary of words is generated from combinations of the ordered sequence of symbols based on a frequency at which combinations of symbols appear in the ordered sequence of symbols. A plurality of phrases is generated based an ordered sequence of words from the dictionary observed in the ordered sequence of symbols based on a frequency by which combinations of words in ordered sequence of words appear relative to one another.

Abstract: An apparatus on a vehicle comprises one or more sensors, one or more nozzles that output fluid to clean the respective one or more sensors, and a compressor that generates fluid such as compressed air. The compressor is in fluid communication with the one or more nozzles. The apparatus further comprises one or more processors, and a memory storing instructions that, when executed by the one or more processors, cause the system to determine a current velocity of the vehicle and control an operation of the compressor based on the current velocity of the vehicle.

Abstract: Computer-based natural language understanding of input and output for a computer interlocutor is improved using a method of classifying conversation segments from transcribed conversations. The improvement includes one or more methods of splitting transcribed conversations into groups related to a conversation ontology using metadata; identifying dominant paths of conversational behavior by counting the frequency of occurrences of the behavior for a given path; creating a conversation model comprising conversation behaviors, metadata, and dominant paths; and using the conversation model to assign a probability score for a matched input to the computer interlocutor or a generated output from the computer interlocutor.

Abstract: An initial environment navigation model for a physical environment may be determined based on sensor data collected from a mobile enrollment device. The sensor data may include data collected from a first one or more cameras at the mobile enrollment device. The initial environment navigation model may be sent to a robot via a communication interface. The robot may be instructed to autonomously navigate the physical environment based on the initial environment navigation model and additional sensor data collected by the robot. An updated environment navigation model for the physical environment may be determined based on the initial environment navigation model and the additional sensor data.

Abstract: A system 100 for automatically generating a field of a radiology report includes a set of one or more models. A method for automatically generating a field of a radiology report includes: receiving a radiologist identifier (radiologist ID); receiving a set of finding inputs; determining a context of each of the set of finding inputs; determining text associated with a portion or all of the radiology report based on the context and the radiologist style; and inserting the text into the report.

Abstract: An audio track capturing a conversation between an interviewer and an interviewee during an interview may be segmented, by executing a speaker identification engine, into a plurality of audio segments each being tagged as associated with one of the interviewer or the interviewee. A speech recognition and natural language processing (NLP) engine may be applied to the audio track to determine attributes associated with audio segments being tagged to the interviewer. The attributes may comprise timing parameters associated with the audio segment and a text content of the audio segment. A rule-based analysis engine may be executed, based on the attributes associated with audio segments being tagged to the interviewer, to determine whether the interviewer conducts the interview in compliance with predetermined rules. Responsive to determining that the interviewer does not conduct the interview in compliance with the predetermined rules, generating a notice regarding the non-compliance of the interview.

Abstract: One or more simulated capture paths through a physical environment may be determined for a robot based on an environment navigation model of the physical environment. A plurality of simulated object parameter values may be determined for an object type. Simulated sensor data for a plurality of simulated instances of the object type may be determined based on the one or more simulated capture paths, the environment navigation model, and the simulated object parameter values. An object recognition model to recognize an object corresponding with the object type based on the simulated sensor data.

Abstract: A sales system includes workflows defined for a sales opportunity with one or more stages in each of the workflows. Among the workflows, workflows for budget, interest, time, and human notions of deal progress may provide a dimensional view of deal progress and intervention strategy. The sales system determines recommended actions to progress the sales opportunity in response to a state of the sales opportunity in the one or more stages. The sales system further determines a deal score, probability of success, or value based on progress within the workflows and system variables. The behaviour of the recommendation system is determined by configured rules that specify how to simulate scenarios, as in a Markov decision process setting. An artificial intelligence component infers the optimal action recommendations from the simulation. A sales representative may track their progress in closing their deal through a deal score, as in a game.

Abstract: Improved calibration of a vehicle sensor based on static objects detected within an environment being traversed by the vehicle is disclosed. A first sensor such as a LiDAR can be calibrated to a global coordinate system via a second pre-calibrated sensor such as a GPS IMU. A static object present in the environment is detected such as signage. A type of the detected object is determined from static map data. Point cloud data representative of the static object is captured by the first sensor and a first transformation matrix for performing a transformation from a local coordinate system of the first sensor to a local coordinate system of the second sensor is iteratively redetermined until a desired calibration accuracy is achieved. Transformation to the global coordinate system is then achieved via application of the first transformation matrix followed by a second known transformation matrix.

Abstract: First and second sequenced outputs are accessed. The sequenced outputs contain variants occurring at different carriers and at different carrier positions. Hashes are generated over a selected pattern length of positions for those carrier positions that are shared between the sequenced outputs to produce window hashes for base patterns in first and second sequences. Each sequence is based on the shared carrier positions and the respective sequenced output. The window hashes are non-unique. Window hashes that occur less than a ceiling number times are selected. The selected window hashes are compared between the sequences on a starting position basis such that selected window hashes for base patterns having same start positions in the sequenced outputs are compared. Common window hashes are identified between the sequences based on the comparing. A similarity measure is determined between the sequences based on the common window hashes.

Abstract: A method for processing an electronic image including receiving, by a viewer, the electronic image and a FOV (field of view), wherein the FOV includes at least one coordinate, at least one dimension, and a magnification factor, loading, by the viewer, a plurality of tiles within the FOV, determining, by the viewer, a state of the plurality of tiles in a cache, and in response to determining that the state of the plurality of tiles in the cache is a fully loaded state, rendering, by the viewer, the plurality of tiles to a display.

Abstract: A system for cardiac signal processing, preferably including any or all of: a data collection device, a set of computing and/or processing subsystems, a set of algorithms and/or models, and/or a set of output devices. A method for cardiac signal processing, preferably including processing a set of inputs to determine a set of metrics and determining a set of outputs based on the set of metrics and/or a set of supplementary metrics, and optionally including any or all of: receiving a set of inputs, determining a set of supplementary metrics associated with the set of metrics, and/or triggering the set of outputs.

Abstract: A method for standardized model interaction can include: determining a model composition, receiving an input, converting the input into a standard object, converting the standard input object into a model-specific input (MSI) object, executing the model using the MSI object, converting the output from the model-specific output (MSO) object to a standard object, repeating previous steps for each successive model within the model composition, and providing a final model output.

Abstract: An example method includes receiving a scenario that includes scenario road portions and scenario hazards. Human driver performance metrics based on driving performance of human drivers on first road portions at least generally similar to the scenario road portions when the human drivers encountered first hazards at least generally similar to the scenario are received. Autonomous vehicle performance metrics based on autonomous vehicles driving on second road portions at least generally similar to the scenario road portions and encountering second hazards at least generally similar to the scenario are received. A scenario autonomous vehicle performance assessment based on the human driver performance metrics and the autonomous vehicle performance metrics are generated and provided.

Abstract: The present disclosure relates to systems and methods for filtering electronic activities. The method includes identifying an electronic activity. The method includes parsing the electronic activity to identify one or more electronic accounts in the electronic activity. The method includes determining, responsive to parsing the electronic activity, that the electronic activity is associated with an electronic account of the one or more electronic accounts. The method includes selecting, based on the electronic account, one or more filtering policies associated with the data source provider to apply to the electronic activity. The method includes determining, by applying the selected one or more filtering policies to the electronic activity, to restrict the electronic activity from further processing based on the electronic activity satisfying at least one of the selected one or more filtering policies. The method includes restricting, the electronic activity from further processing.