Abstract: A method includes obtaining a video sequence having multiple video frames and audio data. The method also includes extracting video features associated with at least one face in the video frames and audio features associated with the audio data. The method further includes processing the video features and the audio features using a trained machine learning model. The trained machine learning model performs a multi-tiered fusion of the video features and different subsets of the audio features in order to identify at least one emotion expressed by at least one person in the video sequence. The multi-tiered fusion of the video features and the audio features may include (i) a first fusion of the video features and a first subset of the audio features and (ii) a second fusion of processed features and a second subset of the audio features, where the processed features are based on the first fusion.

Abstract: A root cause associated with an anomalous event in a device is determined. A method includes retrieving one or more event records associated with the device from a database. The method further includes determining a risk category associated with the one or more event records based on information present in the one or more event records. The risk category indicates a risk associated with a functioning of the device. Additionally, the method includes determining a priority associated with each of the one or more event records based on a baseline associated with the one or more event records. The baseline is defined based on a set of events that occur during a normal functioning of the device. The method includes determining the root cause associated with the anomalous event in the device based on the risk category and the priority associated with the event records.

Abstract: A method implemented by one or more computing systems includes accessing a set of content objects, in which each content object of the set of content objects is pre-labeled with concepts of a plurality of concepts organized according to a hierarchical relationship. The method further includes training, by a machine-learning model, a classification model for classifying content objects within the set of content objects. Training the classification model includes determining, for each object, a plurality of classification values corresponding to the plurality of concepts, calculating a loss for each of the plurality of classification values based on the pre-labeled concepts associated with the object, utilizing a hierarchical constraint loss function to calculate a maximum loss based on the calculated loss for each of the plurality of classification values, and updating the classification model based on the hierarchical constraint loss function until the maximum loss satisfies a predetermined criterion.

Abstract: System and method for computing critical data of multiple entities is provided. Elements, from a set of predetermined elements, are assigned to sentences of a first dataset associated with an entity of an industry type. Predetermined elements characterize reputation of entities of various industries. Features determined from sentences of first dataset are categorized into groups. Features of a new dataset of a different entity of said industry type are matched with grouped features and same groups are allotted thereto. Classification rules are applied on matched features. Elements are assigned to sentences of new dataset based on allotted groups, which elements are same as those assigned to first dataset. Reputation scores are generated for said entities by determining positive and negative sentiments from the first and new dataset. Steps of assigning elements and grouping are repeated for datasets associated with entities of different industry types for creating a taxonomy for them.

Abstract: System and method for computing critical data of multiple entities is provided. Elements, from a set of predetermined elements, are assigned to sentences of a first dataset associated with an entity of an industry type. Predetermined elements characterize reputation of entities of various industries. Features determined from sentences of first dataset are categorized into groups. Features of a new dataset of a different entity of said industry type are matched with grouped features and same groups are allotted thereto. Classification rules are applied on matched features. Elements are assigned to sentences of new dataset based on allotted groups, which elements are same as those assigned to first dataset. Reputation scores are generated for said entities by determining positive and negative sentiments from the first and new dataset. Steps of assigning elements and grouping are repeated for datasets associated with entities of different industry types for creating a taxonomy for them.