Described are platforms, systems, and methods to discover relationships among equipment in automated industrial or commercial environments by cycling each individual piece of equipment while observing sensors in all other equipment in order to measure how each part reacts to each other part. The platforms, systems, and methods identify a plurality of data sources associated with an automation environment; issue one or more commands to cycle a current data source in the a plurality of data sources; monitor the automation environment for events or state changes in the data sources; detect one or more events or one or more state changes in one or more other data sources in the a plurality of data sources; and determine one or more relationships between the current data source and the one or more other data sources.

Embodiments are disclosed for user posture transition detection and classification using a linked biomechanical model. In an embodiment, a method comprises: obtaining motion data from a headset worn by a user; selecting features of a linked biomechanical model based on a current posture state; determining at least one probability that a posture transition occurred based on an output of a classifier, where the output of the classifier is based on the selected features and the motion data; determining a posture transition based on the at least one probability; and performing at least one action based on detection of the posture transition.

Electronic message text classification framework selection is described. An incoming electronic message is classified using a current text classification framework. A classification of the electronic message by the current text classification framework is scored. A cost of re-training the current text classification is compared against a cost of switching to a different text classification framework. One of multiple text classification frameworks, which includes the current text classification framework and other text classification frameworks, is selected based on the score of the classification by the current text classification framework and a result of the comparison.

In various examples, high beam control for vehicles may be automated using a deep neural network (DNN) that processes sensor data received from vehicle sensors. The DNN may process the sensor data to output pixel-level semantic segmentation masks in order to differentiate actionable objects (e.g., vehicles with front or back lights lit, bicyclists, or pedestrians) from other objects (e.g., parked vehicles). Resulting segmentation masks output by the DNN(s), when combined with one or more post processing steps, may be used to generate masks for automated high beam on/off activation and/or dimming or shading—thereby providing additional illumination of an environment for the driver while controlling downstream effects of high beam glare for active vehicles.

A medical knowledge database including medical knowledge information, medical diagnoses, and medical treatments, is used for machine learning for collaborative medical data metrics. Medical data is collected from a plurality of clinicians serving a first plurality of patients and assembling a medical knowledge database that includes medical knowledge information, medical diagnoses, and medical treatments. The medical knowledge database is a function of demographics and comprises a medical probabilistic rules graph. The medical knowledge database is augmented based on further medical data collected from a second plurality of clinicians. The further medical data is based on individual patient treatment outcomes collected by the second plurality of clinicians. Medical data from a further patient is applied to the medical probabilistic rules graph. A medical diagnosis is provided, based on the medical data applied from a further patient to the rules graph. The medical diagnosis is used to institute a treatment plan.

Two models are first trained and then test images are applied to the two trained models in an effort to detect signature forgeries. The first model is trained with pairs of signature images and the resultant trained model is capable of detecting blind forgeries. The second model is trained with triplets of signature images and is capable of detecting skilled signature forgeries. After the two models are trained, test images are applied to the models and determinations are made as to whether a blind or skilled forgery is present.

A method and apparatus are provided that analyzing sequence-to-sequence data, such as sequence-to-sequence speech data or sequence-to-sequence machine translation data for example, by minimum Bayes risk (MBR) training a sequence-to-sequence model and within introduction of applications of softmax smoothing to an N-best generation of the MBR training of the sequence-to-sequence model.

Among other things, techniques are described for spatially and temporally consistent ground modelling with information fusion. A vehicle location and orientation is obtained and an instantaneous ground height estimation is obtained for anchor cells of a grid map of the vehicle based on the obtained vehicle location and orientation. A pseudo ground height associated with non-anchor cells of the grid map is computed. A Bayesian filtering framework is used to generate a final ground height estimate for the anchor cells and the non-anchor cells based on the instantaneous ground height estimation for the anchor cells, the pseudo ground height associated with non-anchor cells, an estimated ground height from a previous timestamp, or any combinations thereof. The vehicle operates according to the final ground height estimate.

A system, method and apparatus for detecting facial expressions according to EMG signals.

In some implementations, an event timeline that includes one or more interactions between a customer and a supplier may be determined. A starting value may be assigned to individual events in the event timeline. A sub-sequence comprising a portion of the event timeline that includes at least one reference event may be selected. A classifier may be used to determine a previous relative value for a previous event that occurred before the reference event and to determine a next relative value for a next event that occurred after the reference event until all events in the event timeline have been processed. The events in the event timeline may be traversed and a monetized value index assigned to individual events in the event timeline.