Abstract: Systems and techniques are provided for generating depth information. For example, a process can include obtaining a first feature volume including visual features corresponding to each respective frame included in a first set of frames. A first query generator network can generate reconstruction features associated with a reconstructed feature volume corresponding to the first feature volume. Based on the first feature volume, a second query generator network can generate motion features associated with predicted future motion corresponding to the first feature volume. An initial depth prediction can be generated for each respective frame based on cross-attention between features of a depth prediction decoder, the reconstruction features, and the motion features. A refined depth prediction can be generated for each respective based on cross-attention between the initial depth prediction, the reconstruction features, and the motion features.

Abstract: Disclosed herein are pharmaceutical compositions in the form of a suspension for oral delivery. Some embodiments provide a pharmaceutical composition in the form of a suspension for oral delivery comprising an active pharmaceutical ingredient; water; a suspending agent; a buffering agent; and one or more of a wetting agent and a binder/filler. In some embodiments, the active pharmaceutical ingredient is selected from quetiapine, sildenafil, tadalafil, cinacalcet, ticagrelor, mycophenolate, aprepitant, zonisamide, and primidone.

Abstract: Systems and techniques are provided for generating depth information from one or more images. For example, a process can include obtaining a first depth map corresponding to an input comprising an image of the one or more images and a sparse depth measurement. A three-dimensional (3D) point cloud can be generated based on the first depth map and multi-scale visual features of the input, wherein the 3D point cloud includes a plurality of 3D point features uplifted from the multi-scale visual features. At least a portion of the plurality of 3D point features can be processed using one or more self-attention layers to generate refined 3D point features. A two-dimensional (2D) projection of the refined 3D point features can be generated and a second depth map can be generated based on the 2D projection of the refined 3D point features.

Abstract: An integrated circuit (IC) includes one or more active transistors and multiple series-coupled dummy transistors. The dummy transistors are coupled between two active transistors and/or at the ends of each active transistor. When the dummy transistors are coupled between two active transistors, apart from two conductive regions that are coupled to two active transistors, each remaining conductive region of the dummy transistors is maintained in a floating state to control a leakage current between the two active transistors. Similarly, when the dummy transistors are coupled at an end of one active transistor, apart from one conductive region that is coupled to the active transistor, each remaining conductive region of the dummy transistors is maintained in the floating state to control a leakage current between the active transistor and the dummy transistors.

Abstract: Stateless playback is performed by receiving a request from a media playback device, the request requesting a playback position associated with a collaborative media consumption session. Upon receiving the request, a playback position for the collaborative media consumption session is computed. The playback position is then communicated to the media playback device.

Abstract: At least one non-transitory machine-readable medium includes machine-readable instructions to cause at least one processor circuit to at least: receive a first event message from an event log, the first event message including (i) a first identifier of a first event for a first process and (ii) a first timestamp of the first event; determine, based on the first identifier, a type of the first event to be a client call finish; determine the first identifier is active; record the first identifier, the first timestamp, and the first event in a memory; determine that at least one of a client call interface registration event or a client call endpoint registration event that has the first identifier was received; reattribute the first process; and remove the first identifier, the first timestamp, and the first event from the memory.

Abstract: Methods and apparatuses directed to memory fault detection mechanisms within die architectures. In some examples, a die package includes decoder logic that receives multiple data words, and a first error correcting code for each of the data words. The decoder logic generates, for each of the data words, a second error correcting code based on a corresponding one of the data words. Further, the decoder logic generates, for each of the data words, an error status based on the first error correcting code and the second error correcting code that corresponds to each of the data words. The die package also includes error generation logic that receives the error status for the data words from the decoder logic, and generates error data based on a combination of the error statuses for the data words. The error generation logic can store the error data in a memory device.

Abstract: The present invention relates to Human Immunodeficiency Virus (HIV) prevention and treatment. In particular, the invention relates to a pharmaceutical composition comprising: cabotegravir; a wetting agent; a stabilizer; and a tonicity adjuster; wherein cabotegravir is present in the form of particles having a mass median diameter (X50) of between (and including) 2.5 ?m and 10 ?m.

Abstract: A fluid adhesive dispenser device includes a housing having a passageway through the housing and a port at a distal end of the passageway. The dispenser device also includes a container at least partially within the housing configured to contain a single-use amount of the fluid adhesive. The container is configured to be opened allowing the fluid adhesive contained therein to be actively expelled from the container through the passageway and the port of the housing. A method of manufacture of the fluid adhesive dispenser device includes providing the dispenser device filled with the single-use amount of the fluid adhesive and sterilizing the filled device.

Abstract: A system includes computer-readable media configured to store a plurality of objects representing intent graph models of a network, and processing circuitry coupled to the computer-readable media. The processing circuitry is configured to receive a request indicating a requested time, determine one or more first objects of the plurality of objects, the first objects storing an intent graph model associated with a first time, the first time different from the requested time, determine one or more second objects of the plurality of objects, the second objects storing difference information indicating one or more changes to the intent graph model associated with the first time that occurred after the first time, apply the changes to the intent graph model associated with the first time to generate an intent graph model associated with the requested time, and output an indication of the intent graph model associated with the requested time.

Abstract: A processor-implemented method for image or text processing includes receiving, by an artificial neural network (ANN) model, a set of tokens corresponding to an input. A token interaction block of the ANN model processes the set of tokens according to each channel of the input to generate a spatial mixture of a set of features for each channel of the input. A feed forward network block of the ANN model generates a mixture of channel features based on the spatial mixture of the set of features for each channel of the input. An attention block of the ANN model determines a set of attended features of the mixture of channel features according to a set of attention weights. In turn, the ANN model generates an inference based on the set of attend features of the mixture of channel features.

Abstract: To automatically onboard network functions to a credential vault, a orchestration processor actuates establishment of an cluster account for a network cluster, and actuates a cluster configuration of a processor of the vault to enable authentication of a network cluster. For each of a plurality of network functions associated with the network cluster, the orchestration processor generates an identifier, sets values for parameters of an initialization parameter set, actuates assignment of access permissions for a code address on a memory of the vault, actuates assignment of elevated access permissions for a credential address on the vault memory, and actuates association of the network function with a cluster account of the network cluster. The vault memory thereby defines credential addresses each corresponding to a respective network function.

Abstract: A machine learning based computing method for computing optimal time for sending email, is disclosed. The machine learning based computing method includes steps of: receiving inputs from second users; extracting data associated with the first users and the second users from databases, based on the inputs received from the second users; computing electronic mail feature scores based on the extracted data associated with the first users and the second users, for each specified interval of contact prediction window; computing first user electronic mail scores for each specified interval of the contact prediction window based on the electronic mail feature scores for each specified interval of the contact prediction window, using a machine learning model; and computing at least one of: the optimal times and a prioritized list of the optimal times by ranking each specified interval of the contact prediction window associated with the first user electronic mail scores.

Abstract: Disclosed herein are pharmaceutical compositions in the form of a suspension for oral delivery. Some embodiments provide a pharmaceutical composition in the form of a suspension for oral delivery comprising an active pharmaceutical ingredient; water; a suspending agent; a buffering agent; and one or more of a wetting agent and a binder/filler. In some embodiments, the active pharmaceutical ingredient is selected from quetiapine, sildenafil, tadalafil, cinacalcet, ticagrelor, mycophenolate, aprepitant, zonisamide, and primidone.

Abstract: The present disclosure describes compositions and methods for preparing membrane protein nanosheets and two-dimensional crystals. In particular, the methods employ a solvent. A mixture of a polymer and a membrane protein is solubilized in the solvent, applied to a substrate, and subsequently dried to form the nanosheet or two-dimensional crystal. Applicants have surprisingly found that the membrane proteins maintain their structure when exposed to solvents during the short processing time utilized.

Abstract: A vascular access device (VAD) monitoring system includes a VAD with a catheter tube that is subcutaneously disposed and incorporates a fluorescent dye. An optical imaging system is also provided, comprising a light source emitting infrared (IR) excitation light and a camera capable of detecting signal light emitted by the fluorescent dye upon exposure to the excitation light. This enables real-time monitoring of the VAD, allowing for the detection of any potential complications or issues related to the vascular access device. The combination of the catheter tube with the fluorescent dye and the optical imaging system provides a non-invasive and efficient means of monitoring the VAD, enhancing patient safety and improving overall healthcare outcomes.

Abstract: A system and method for a risk management visualization system having a computer comprising a processor and memory and configured to model a supply chain network as a supply chain planning problem, one or more key process indicators (KPIs) of the supply chain planning problem is based, at least in part, on the one or more input variables, model an impact on the one or more KPIs from each of the one or more input variables at a selected confidence interval using a Bayesian optimization process comprising an exploration phase and a learning phase, and display a visualization of the risk profile for the one or more KPIs, the visualization indicating a probability that an actual KPI value differs from a predicted KPI value.

Abstract: A system and method for a risk management visualization system having a computer comprising a processor and memory and configured to model a supply chain network as a supply chain planning problem, one or more key process indicators (KPIs) of the supply chain planning problem is based, at least in part, on the one or more input variables, model an impact on the one or more KPIs from each of the one or more input variables at a selected confidence interval using a Bayesian optimization process comprising an exploration phase and a learning phase, and display a visualization of the risk profile for the one or more KPIs, the visualization indicating a probability that an actual KPI value differs from a predicted KPI value.

Abstract: An apparatus includes a memory that stores instructions; and a processing unit that executes the instructions to identify a created process, to receive a notification of a first event for an ancestor process and a notification for a second event for the created process, the notification of the first event indicating a first ActivityID and a first ID, the notification of the second event indicating a second ActivityID and a second ID, the first ID being different from the second ID, to perform a first determination that the created process was created by a component object model (COM) call, at least in part based on the second ID, and to perform a second determination that the ancestor process indirectly created the created process, at least in part based on the first and second ActivityIDs and the first determination.

Abstract: A system and method for a risk management visualization system having a computer comprising a processor and memory and configured to model a supply chain network as a supply chain planning problem, one or more key process indicators (KPIs) of the supply chain planning problem is based, at least in part, on the one or more input variables, model an impact on the one or more KPIs from each of the one or more input variables at a selected confidence interval using a Bayesian optimization process comprising an exploration phase and a learning phase, and display a visualization of the risk profile for the one or more KPIs, the visualization indicating a probability that an actual KPI value differs from a predicted KPI value.