A reinforcement learning architecture for facilitating reinforcement learning in connection with operation of an external real-time system that includes a plurality of devices operating in a real-world environment. The reinforcement learning architecture includes a plurality of communicators, a task manager, and a reinforcement learning agent that interact with each other to effectuate a policy for achieving a defined objective in the real-world environment. Each of the communicators receives sensory data from a corresponding device and the task manager generates a joint state vector based on the sensory data. The reinforcement learning agent generates, based on the joint state vector, a joint action vector, which the task manager parses into a plurality of actuation commands. The communicators transmit the actuation commands to the plurality of devices in the real-world environment.

Systems and methods for collecting data from multiple sources. A server receives a data request for desired content from a requestor. The server generates a request for data objects containing that content and displays that request to at least one potential data provider. The at least one data provider provides a data object in response to the request. In some embodiments, the data provider may label the provided data object. In further embodiments, labels may replace sensitive personal information on the data. The data objects may be obtained by the data providers using their personal portable computing devices. In some embodiments, data providers may upload data objects that are not directed to a specific request. These data objects may then be associated with later data requests.

Certain aspects of the present disclosure provide techniques for automatically detecting and classifying tumor regions in a tissue slide. The method generally includes obtaining a digitized tissue slide from a tissue slide database and determining, based on output from a tissue classification module, a type of tissue of shown in the digitized tissue slide. The method further includes determining, based on output from a tumor classification model for the type of tissue, a region of interest (ROI) of the digitized tissue slide and generating a classified slide showing the ROI of the digitized tissue slide and an estimated diameter of the ROI. The method further includes displaying on an image display unit, the classified slide and user interface (UI) elements enabling a pathologist to enter input related to the classified slide.

In some embodiments, a first plurality of electron-microscope images for respective instances of a semiconductor structure is obtained from a first source. The electron-microscope images of the first plurality show different values of one or more semiconductor-fabrication parameters. A model is trained that specifies a relationship between the first plurality of electron-microscope images and the values of the one or more semiconductor-fabrication parameters. A second plurality of electron-microscope images for respective instances of the semiconductor structure on one or more semiconductor wafers is collected. The one or more semiconductor wafers are distinct from the first source. Values of the one or more semiconductor-fabrication parameters for the second plurality of electron-microscope images are predicted using the model.

Personalized video interjections based on a learner model and a learning objective. A method for adding interjections to a video includes analyzing the content of a plurality of videos based on a set of learning objectives, selecting a video based on a learning objective, determining types of video interjections using an analytics engine that compares a learner model and the learning objective, determining a location for the video interjections using the analytics engine, generating a video interjection for each video interjection type and inserting the video interjections into the video at the determined locations.

System/method identifying a defined object (e.g., hazard): a sensor detecting and defining a digital representation of an object; a processor (connected to the sensor) which executes two techniques to identify a signature of the defined object; a memory (connected to the processor) storing reference data relating to two signatures derived, respectively, by the two techniques; responsive to the processor receiving the digital representation from the sensor, the processor executes the two techniques, each technique assessing the digital representation to identify any signature candidate defined by the object, derive feature data from each identified signature candidate, compare the feature data to the reference data, and derive a likelihood value of the signature candidate corresponding with the respective signature; combining likelihood values to derive a composite likelihood value and thus determine whether the object in the digital representation is the defined object.

An artificial intelligence apparatus for recognizing a user includes a camera, and a process configured to receive, via the camera, image data including a recognition target object, generate recognition information corresponding to the recognition target object from the received image data, calculate a confidence level of the generated recognition information, determine whether the calculated confidence level is greater than a reference value, if the calculated confidence level is greater than the reference value, perform a control corresponding to the generated recognition information, and if the calculated confidence level is not greater than the reference value, provide a feedback for the object recognition.

Provided are systems, methods, and media for teaching generalization of an object. An example method includes obtaining a set of traits of an object recognized by a person in an input image, in which a subset of traits are traits fixated on by the person when recognizing the object in the input image. Executing a machine learning algorithm to generate a set of generalized images of the object. Each generalized image is generated with at least one trait of being modified, in which the set of generalized images are ordered in a sequence based on proximity of each of the generalized images to the input image. Presenting at least a first generalized image to the person in accordance with the sequence. Modifying the order of the generalized images in the sequence in response to detecting from feedback that the person does not recognize the object in the first generalized image.

A method is provided for classifying objects. The method detects objects in one or more images. The method tags each object with multiple features. Each feature describes a specific object attribute and has a range of values to assist with a determination of an overall quality of the one or more images. The method specifies a set of training examples by classifying the overall quality of at least some of the objects as being of an acceptable quality or an unacceptable quality, based on a user's domain knowledge about an application program that takes the objects as inputs. The method constructs a plurality of first-level classifiers using the set of training examples. The method constructs a second-level classifier from outputs of the first-level automatic classifiers. The second-level classifier is for providing a classification for at least some of the objects of either the acceptable quality or the unacceptable quality.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for learning template representation libraries. In one aspect, a method includes obtaining an image depicting a physical environment, where the environment includes a given physical object. When possible, a position of the given object in the environment is inferred based on a template representation library using template matching techniques. In response to determining that the position of the given object in the environment cannot be inferred based on the template representation library using template matching techniques, the template representation library is automatically augmented with new template representations.