Abstract: Disclosed herein are system, method, and computer program product embodiments for machine learning systems to process incoming call-center calls based on inferred themes. The machine learning system extracts a topic and keywords associated with the topic from a plurality of interactive communications and ranks the keywords based on a frequency of occurrence within the plurality of interactive communications. The machine learning systems select an N highest ranked keywords from the plurality of interactive communications, compares the N highest ranked keywords to previously extracted N highest ranked keywords to identify new keywords, and determines, based on new keywords, that an emerging topic has been articulated in the plurality of interactive communications.

Abstract: Aspects of the subject disclosure may include, for example, identifying a first user associated with a first mobile device, and determining a first user type associated with the first user. Further embodiments can include identifying a first identity associated with the first user based on the first user type and identifying a second identity associated with the first user based on the first user type. Additional embodiments can include providing first alternate content to the first mobile device to mask the first identity and providing second alternate content to the first mobile device associated with the second identity. Other embodiments are disclosed.

Abstract: A neural network training apparatus is described which has a network of worker nodes each having a memory storing a subgraph of a neural network to be trained. The apparatus has a control node connected to the network of worker nodes. The control node is configured to send training data instances into the network to trigger parallelized message passing operations which implement a training algorithm which trains the neural network. At least some of the message passing operations asynchronously update parameters of individual subgraphs of the neural network at the individual worker nodes.

Abstract: Implementations of firearm regulation systems may include a firearm safety system (FSS) configured to physically couple to a firearm. The FSS may include a locking mechanism configured to prevent the firearm from firing and a sensor configured to detect an environmental parameter. The firearm regulation system may also include a management system communicatively coupled to the FSS and a site safety system (SSS) communicatively coupled to the management system. The SSS may include one or more lockdown mechanisms.

Abstract: Endoscopic image analysis, endoscopic procedure analysis, and/or component control systems, methods and techniques are disclosed that can analyze images of an endoscopic system and/or affect an endoscopic system to enhance operation, user and patient experience, and usability of image data and other case data.

Abstract: Compounds having three or more alkynyl moieties substituted with an aromatic moiety having one or more of certain substituents are useful in forming underlayers useful in semiconductor manufacturing processes.

Abstract: Technologies are provided for an extensible software-as-a-service (SaaS) computing platform. In some embodiments, a computing system includes computer-executable components comprising multiple core modules that provide a defined service. The multiple core modules include at least one extension point corresponding to a first core module of the multiple core modules, where a first extension point of the at least one extension point has a tenant-specific extension module mapped thereto that customizes the defined service. The computer-executable components also can include an application programming interface (API) corresponding to a defined tenant. The computing system also includes a data model corresponding to the defined tenant. Tenant-specific extension modules mapped to the at least one extension point can customize the API and/or the data model.

Abstract: Aspects of the disclosure relate to voice activity detection (VAD) on wearable and other resource-constrained devices, to classify speech recorded by a microphone of the device as belonging to a wearer of the device versus another speech source. A computing device can include a microphone and an inertial measurement unit (IMU). The wearable device can use signals measured by the IMU for providing motion-tracking features, such as head tracking for augmented reality or virtual reality applications. Aspects of the disclosure provide for a device for leveraging existing data collected for these motion-tracking features for use in VAD. A device can pre-process data streamed from an IMU to use only signals predetermined to be indicative of whether or not a wearer of the device is speaking.

Abstract: The present disclosure relates to a processes and systems for producing fuels from biomass with high carbon conversion efficiency. The processes and systems described herein provide a highly efficient process for producing hydrocarbons from biomass with very low Green House Gas (GHG) emissions using a specific combination of components, process flows, and recycle streams. The processes and systems described herein provide a carbon conversion efficiency greater than 95% with little to no GHG in the flue gas due to the novel arrangement of components and utilizes renewable energy to provide energy to some components. The system reuses water and carbon dioxide produced in the process flows and recycles naphtha and tail gas streams to other units in the system for additional conversion to syngas to produce hydrocarbon-based fuels.

Abstract: Animal waste contains most of the essential nutrients required for plant growth. However, animal waste also contains microbes that convert urea to ammonium ions, nitrification of the ammonium ions to NO2 and NO3 and denitrification of NO2 or NO3 to N2. N2O is produced as an intermediate of nitrification and denitrification processes and N2O is a 300 times more potent greenhouse gas than CO2. The instant invention discloses low-cost, fluid suspension compositions comprising a combination of urease, nitrification and denitrification inhibitors that decrease the loss of nitrogen from animal waste and lessen the biological formation of the greenhouse gas N2O. These compositions comprise nitrogenous phosphoryl compounds and nitrification inhibitors that utilize manufacturing processes for the formation of low cost, fluid suspensions comprising one or more inhibitor particles selected from the group consisting of urease inhibitor particles, b) nitrification inhibitor particles, and c) fused particles.

Abstract: A vehicle having a Lidar system includes a cleaning apparatus that performs a method of cleaning the Lidar system. The cleaning apparatus includes a nozzle for spraying a gas onto a window of the Lidar, a pump for controlling a flow of the gas through the nozzle, and a processor. The processor is configured to control the pump to control the flow of the gas through the nozzle at a variable flow rate.

Abstract: An electric drive system is actively controlled while an electrified vehicle is being towed. The inverter switches are manipulated to preclude exceeding a maximum voltage of a DC bus and also to preclude exceeding a maximum phase current magnitude. During a transition from an inverter off state to a short-circuit state, field-oriented control is utilized.

Abstract: Disclosed herein are system, method, and computer program product embodiments for replicating a remote venue in a local venue. An embodiment operates by receiving original audio. Thereafter, the original stream of audio is modified to produce modified audio based on an audio profile unique to a remote venue smaller than the local venue. The audio profile comprises a virtual representation of the remote venue including (i) a virtual audio source corresponding to the local venue's audio source and configured to produce the original audio and (ii) a virtual reverberation point corresponding to a real-world location in the remote venue and the portion of the local venue. As such, the modified audio is determined based on receipt of the original audio from the reverberation point. Thus, after correlating the modified audio to the local venue's audio source, the modified audio is provided by the audio source to the portion of the local venue.

Abstract: The present disclosure provides for a sandstone-containing oil reservoir additive composition that includes (a) a nonionic wettability alteration agent having Formula (I): and (b) an injection fluid. For Formula (I), R is an alkyl residue from a branched secondary alcohol initiator, where R can have a carbon number of C4 to C10, x is from 3 to 10 and y is from 1 to 20. The injection fluid can be selected from the group consisting of sea water, engineered water, injection brine, produced brine or combinations thereof. The nonionic wettability agent in the sandstone reservoir additive composition can be from 0.01 weight percent (wt %) to 20 wt % based on the total weight of the sandstone reservoir additive composition. The sandstone-containing oil reservoir additive composition can be used in enhance oil recovery operations, such as a waterflooding operation or a carbon dioxide flooding operation.

Abstract: A base station transmits a DCI comprising: an allocation of resource blocks for wireless device transmission of uplink data via a transmission time interval (TTI) of an unlicensed cell; an index indicating a starting position for the wireless device transmission of the uplink data, the index identifying one of pre-configured starting positions; and a field indicating an ending symbol for the wireless device transmission of the uplink data. The ending symbol is one of pre-configured ending symbols. The reception is based on a listen before talk (LBT) procedure indicating a clear channel before the starting position of the subframe.

Abstract: A method includes receiving user input indicating a selection of a subset of two or more of a plurality of audio layers for media content to be provided to a user, each of the subset of audio layers corresponding to one or more audio sources, receiving second user input indicating volume levels for the two or more audio layers of the plurality of audio layers, capturing, based on the first user input, the two or more audio layers of the plurality of audio layers for a media content item to be provided to the user, and enabling audio playback based on the two or more audio layers of the plurality of audio layers and without including other audio layers of the plurality of audio layers, the audio playback reflecting the volume levels indicated by the second user input. The method further includes causing the media content item to be provided to the user using the audio playback reflecting the indicated volume levels.