Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to an artificial neural network training method that does not require extensive training data or time expenditure. The few-shot training model disclosed herein includes attempting to pick up items and, in response to a failed pick up attempt, transferring and generalizing information to similar regions to improve probability of success in future attempts. In some implementations, the training method is used to robotic device for picking items from a bin and perturbing items in a bin. When no picking strategies with high probability of success exist, the robotic device may perturb the contents of the bin to create new available pick-up points. In some implementations, the device may include one or more computer-vision systems.

Abstract: Provided are a resin composition for 3D printer which is low in viscosity, is good in handleability by 3D printers and can give shaped products excellent in flexibility in a broad temperature range, and a method for producing the same, and a cured product of the resin composition. The resin composition for 3D printer of the present invention contains a first monomer, a second monomer, and a photopolymerization initiator, wherein the first monomer is a reaction product of a particular polyether polyol, a polyisocyanate and a particular compound having a (meth)acryloyl group; the total number of moles of the hydroxyl group of the polyether polyol and a group reactive with isocyanate group of the compound having a (meth)acryloyl group is equal to the number of moles of the isocyanate group of the polyisocyanate; and the second monomer is a (meth)acrylic compound.

Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to an artificial neural network training method that does not require extensive training data or time expenditure. The few-shot training model disclosed herein includes attempting to pick up items and, in response to a failed pick up attempt, transferring and generalizing information to similar regions to improve probability of success in future attempts. In some implementations, the training method is used to robotic device for picking items from a bin and perturbing items in a bin. When no picking strategies with high probability of success exist, the robotic device may perturb the contents of the bin to create new available pick-up points. In some implementations, the device may include one or more Computer-vision systems.

Abstract: Systems and methods for allowing a subscriber to opt-out of targeted digital advertisements are provided. In one implementation, a mobile network operator operations support system server receives an input from a subscriber, the input comprising an account number and a stable network-level identifier. The server then causes a message to be sent to the mobile device along with a URL based on the input from the subscriber. The server receives a beacon that is generated when the subscriber visits the website, and sends a message to a mobile analytics platform server indicating a preference on whether the subscriber wants to receive targeted digital advertisements.

Abstract: While multiple-choice questions can be used to objectively assess individuals, presenting individuals with potential responses in a multiple-choice format can unintentionally influence the individual answering the question. Accordingly, the disclosed system provides functionality for users to answer questions using natural language and identifies the potential responses to those questions that are most similar to the natural language responses provided by the users. To assess users who use different dialects, the system includes a universal sentence encoder that recognizes semantic concepts regardless of the dialect used by those users.

Abstract: Provided are a resin composition with low viscosity and a resin cured product. The resin composition comprises a first monomer and a second monomer, wherein the proportion of the first monomer with respect to the total mass of the first monomer and the second monomer is 50 to 98% by mass, and the resin cured product is of the resin composition.

Abstract: Various embodiments of the technology described herein generally relate to systems and methods for trajectory optimization with machine learning techniques. More specifically, certain embodiments relate to using neural networks to quickly predict optimized robotic arm trajectories in a variety of scenarios. Systems and methods described herein use deep neural networks to quickly predict optimized robotic arm trajectories according to certain constraints. Optimization, in accordance with some embodiments of the present technology, may include optimizing trajectory geometry and dynamics while satisfying a number of constraints, including staying collision-free, and minimizing the time it takes to complete the task.

Abstract: Various embodiments of the technology described herein generally relate to systems and methods for trajectory optimization with machine learning techniques. More specifically, certain embodiments relate to using neural networks to quickly predict optimized robotic arm trajectories in a variety of scenarios. Systems and methods described herein use deep neural networks to quickly predict optimized robotic arm trajectories according to certain constraints. Optimization, in accordance with some embodiments of the present technology, may include optimizing trajectory geometry and dynamics while satisfying a number of constraints, including staying collision-free and minimizing the time it takes to complete the task.

Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to modeling uncertainty in neural network segmentation predictions of imaged scenes having a plurality of objects. In some embodiments, a computer vision system for guiding robotic picking utilizes a method for uncertainty modeling that comprises receiving one or more images of a scene comprising a plurality of distinct objects, generating a plurality of segmentation predictions each comprising one or more object masks, identifying a predefined confidence requirement, wherein the confidence requirement identifies a minimum amount of required agreement for a region, and outputting one or more object masks based on the confidence requirement. The systems and methods disclosed herein leverage the use of a plurality of hypotheses to create a distribution of possible segmentation outcomes in order model uncertainty associated with image segmentation.

Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to modeling uncertainty in neural network predictions using bounding box predictions for imaged objects. In some embodiments, a computer vision system for guiding robotic picking utilizes a method for uncertainty modeling that comprises identifying a three-dimensional object in one or more images of a scene, wherein at least one side of the 3D object is not visible to the computer vision system. The method further comprises predicting a plurality of volumes that comprise the object, wherein each volume of the plurality of volumes comprises at least a portion of the object. From the plurality of volumes, a confidence level may be determined for each volume, wherein the confidence level represents a likelihood that the volume contains the entire object.

Abstract: Various embodiments of the present technology generally relate to robotic devices, computer vision, and artificial intelligence. More specifically, some embodiments relate to object tracking using neural networks and computer vision systems. In some embodiments, a computer vision system for object tracking captures one or more images of a first scene, wherein the first scene corresponds to a first location, identifies a distinct object in the first scene based on the one or more first images, directs a robotic device to move the distinct object from the first location to a second location, captures one or more second images of a second scene, wherein the second scene corresponds to the second location, and determines if the distinct objects is in the second scene based on the one or more second images.

Abstract: Various embodiments of the present technology generally relate to robotic devices, computer-vision systems, and artificial intelligence. More specifically, some embodiments relate to a computer-vision system for robotic devices. In some implementations, a computer-vision is coupled to a robotic arm for picking items from a bin and perturbing items in a bin. A computer-vision system, in accordance with the present technology, may use at least two two-dimensional (2D) images to generate three-dimensional (3D) information about the bin and items in the bin. A method of training artificial neural networks to generate 3D information from 2D images includes obtaining ground truth data for a scene, obtaining at least two images of the scene, and providing the ground truth data and the at least two images to the artificial neural network configured to generate a depth map from the images based on the ground truth data.

Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to an artificial neural network training method that does not require extensive training data or time expenditure. The few-shot training model disclosed herein includes attempting to pick up items and, in response to a failed pick up attempt, transferring and generalizing information to similar regions to improve probability of success in future attempts. In some implementations, the training method is used to robotic device for picking items from a bin and perturbing items in a bin. When no picking strategies with high probability of success exist, the robotic device may perturb the contents of the bin to create new available pick-up points. In some implementations, the device may include one or more Computer-vision systems.

Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to a robotic device for picking items from a bin and perturbing items in a bin. In some implementations, the device may include one or more computer-vision systems. A computer-vision system, in accordance with the present technology, may use at least two two-dimensional images to generate three-dimensional (3D) information about the bin and items in the bin. Based on the 3D information, a strategy for picking up items from the bin is determined. When no strategies with high probability of success exist, the robotic device may perturb the contents of the bin to create new available pick-up points and re-attempt to pick up an item.

Abstract: Software for an online content service obtains a plurality of events chronologically generated by a plurality of users of an online content service during a specified period of time. The software identifies any content items associated with each event and annotates each of the content items with (a) a plurality of metadata attributes associated with the content item and (b) a plurality of metadata attributes associated with the online content service. The software sorts the events based on user and based on content identifier and orders the sorted events based on timestamp. The software determines the events that make up a content session for the specific content item and the specific user, using the ordered events for the specific content item and a look-back time period and a look-ahead time period. Then the software generates an analytic based at least in part on the content session.

Abstract: Systems, methods and computer program products for measuring conversion in online digital advertising solve different technical challenges in measuring digital advertising conversion across browsers, applications, domains and devices without deploying cookies.

Abstract: Systems and methods for allowing a subscriber to opt-out of targeted digital advertisements are provided. In one implementation, a mobile network operator operations support system server receives an input from a subscriber, the input comprising an account number and a stable network-level identifier. The server then causes a message to be sent to the mobile device along with a URL based on the input from the subscriber. The server receives a beacon that is generated when the subscriber visits the website, and sends a message to a mobile analytics platform server indicating a preference on whether the subscriber wants to receive targeted digital advertisements.

Abstract: Software for an online content service obtains a plurality of events chronologically generated by a plurality of users of an online content service during a specified period of time. The software identifies any content items associated with each event and annotates each of the content items with (a) a plurality of metadata attributes associated with the content item and (b) a plurality of metadata attributes associated with the online content service. The software sorts the events based on user and based on content identifier and orders the sorted events based on timestamp. The software determines the events that make up a content session for the specific content item and the specific user, using the ordered events for the specific content item and a look-back time period and a look-ahead time period. Then the software generates an analytic based at least in part on the content session.

Abstract: A system and method is described wherein multiple users have provisions to author and review contract documents. A computer based system for plurality of parallel contract reviewing is provided. The system comprises a processor unit and a computer readable medium storing instructions executable by the processor unit to perform the step of inserting the contract marks in terms of bookmarks for all the extracted clauses of the documents. The tool reads the documents in terms of bookmarks. A bookmark identifies a location or a selection of text that a user may name and identify for future reference. For example, a user might use a bookmark to identify text that another user wants to revise at a later time. Likewise, for every contract document, the tool automatically places bookmarks so that every section and clause can be identified in terms of different bookmarks.

Abstract: A system can include a server that includes or is associated with a revenue analysis module. The module can be configured to receive user web browsing session data from a data source. The session data may include page view data, page interaction data, and page referral data. The module may also be configured to determine page referral events, revenue generating events, and relationships between the page referral events and the revenue generating events, according to at least part of the session data. The revenue generating events may include events in which a user interaction with the webpage generates revenue. The module may also be configured to determine revenue generated by the revenue generating events per page referral event of the page referral events, according to the relationships between the page referral events and the revenue generating events.