Abstract: Described is a system for object detection that is robust to adversarial attacks. An initial hypothesis of an identity of an object in an input image is generated using a sparse convolutional neural network (CNN) and a distribution aware classifier. A foveated hypothesis verification process is performed for identifying a region of the input image that supports the initial hypothesis. Using a part-based classifier, an identity of a part of the object in the region of the input image is predicted. An attack probability for the predicted identity of the part, and the initial hypothesis is updated based on the predicted identity of the part and the attack probability. The foveated hypothesis verification process and updating of hypotheses is performed until a hypothesis reaches a certainty threshold. The object is labeled based on the hypothesis that reached the certainty threshold.

Abstract: Identifying polyps or lesions in a colon. In some variations, computer-implemented methods for polyp detection may be used in conjunction with an endoscope system to analyze the images captured by the endoscopic system, identify any polyps and/or lesions in a visual scene captured by the endoscopic system, and provide an indication to the practitioner that a polyp and/or lesion has been detected.

Abstract: A nutritional composition that includes a milk fat globule membrane component, at least one human milk oligosaccharide, and at least one additional component chosen from the group consisting of DHA, ARA, Vitamin E, Vitamin C, and sphingomyelin. The nutritional composition is in the form of a ready-to-drink composition, a food supplement, an extruded food bar, or a solid oral dosage form. The nutritional composition is typically administered or provided to adults, children, or infants through consumption.

Abstract: A device receives, from a cloud computing environment, cloud instance information associated with cloud instances in the cloud computing environment, and processes the cloud instance information, with a machine learning model, to determine containers for one or more of the cloud instances and whether cloud instances should be removed from the cloud computing environment. The device causes a first subset of the cloud instances to be removed from the cloud computing environment, based on determining which of the cloud instances should be removed, and causes the containers to be created for a second subset of the cloud instances based on determining the containers. The device receives, from the cloud computing environment, cloud container information associated with the containers created in the cloud computing environment, and causes one or more of the containers to be scaled based on the cloud container information.

Abstract: An approach for encoding/decoding video data including obtaining video data; parsing the obtained video data to obtain partitioning parameters; obtaining a partitioned coding block based on the partitioning parameters; and decoding the obtained video data based on the partitioned coding block. The partitioned coding block may be partitioned into a rectangular partition and at least one non-rectangular partition.

Abstract: An example method of application identification in a virtualized computing system having a cluster of hosts, the hosts including virtualization layers supporting virtual machines (VMs), is described. The method includes: executing, by application analysis software, process discovery agents for the VMs; receiving, at the application analysis software from the process discovery agents, process metadata describing processes executing on the VMs; generating signatures for the processes based on the process metadata; and determining components of an application based on the signatures.

Abstract: A vehicle is provided with a transmission having an inductor assembly. The inductor assembly is mounted within the transmission such that it is directly cooled by transmission fluid through at least one of spraying, splashing and immersion. The transmission includes at least one gear that is configured to, when rotating, transmit torque between an input and output of the transmission and splash fluid onto the inductor assembly to cool the inductor assembly.

Abstract: A media application determines, based on pixels of images or videos from a collection of media items, clusters of media items such that the media items in each cluster have a visual similarity, wherein the collection of media items is associated with a user account. The media application selects a subset of the clusters of media from corresponding clusters of media items based on the media items in each cluster having a visual similarity within a range of threshold similarity values. The media application causes a user interface to be displayed that includes the subset of the clusters of media.

Abstract: A distributed studio management system includes a computing system having a computer readable memory and a processor. The computing system is associated with a first production facility that is remote to a second production facility. The processor is configured to perform the operations of receiving a set of device commands having commands targeting first and second devices in the first production facility and where the set of device commands are related to the second production facility, enqueuing a first command from the device commands into a first command queue assigned to the first device and a second command from the device commands into a second command queue assigned to the second device according to each device command's synchronized execution times, and causing the first and second devices to execute commands from their respective command queues relative to each other according to the commands' respective synchronized execution times.

Abstract: Systems and methods are disclosed for attributing web traffic to an advertising spot. The method may include receiving traffic data for a web page from a server associated with an advertiser and receiving, from a log provider, a log of a plurality of advertising spots related to the advertiser. A duration of time as a peak may be designated to identify the amount of traffic that is attributable to the one of the plurality of advertising spots.

Abstract: A server application can serve a web application to a client device as a client web application from a web application host. The server application can be a secure HTTPS website. The web application on the client device can provide a token over a local network using a HTTP request to a target device on the local network. The token is associated with a secure area of a remote repository. Having received the one or more tokens, the target device can indicate that a link has been established to the secure area by writing to or reading from the secure area as the server application informs the client web application executing in a browser at a client device of the read/write. The target device and client web application can then communicate via the secure area, for example, to store or retrieve data associated with a user.

Abstract: Methods and systems for automated construction of an anonymized facial recognition library are disclosed. A camera of a client device may capture a first plurality of images of faces of members of a panel of viewers of media content presented on a content presentation device collocated with the client device during viewing sessions. A first machine learning (ML) model may be applied to the first plurality to generate a second plurality of feature vectors, each associated with a different one of the images. One or more clusters of feature vectors of the second plurality may be computationally determined within a vector space of the feature vectors. A respective centroid feature vector may be determined for each respective cluster, and assigned a unique ID. A respective association between each cluster ID and a respective name ID may be determined based on panel-member information received at the client device.

Abstract: An alternating stack of first material layers and second material layers is formed over a substrate. A hard mask layer is formed over the alternating stack. Optionally, an additional hard mask layer can be formed over the hard mask layer. A photoresist layer is applied and patterned, and cavities are formed in the hard mask layer by performing a first anisotropic etch process that transfers a pattern of the openings in the photoresist layer through the hard mask layer. Via openings are formed through an upper portion of the alternating stack by performing a second anisotropic etch process. A cladding liner can be optionally formed on sidewalls of the cavities in the hard mask layer. The via openings can be vertically extend through all layers within the alternating stack by performing a third anisotropic etch process.

Abstract: Aspects of the disclosure involve systems and methods for utilizing Virtual Local Area Network separation in a connection, which may be a single connection, between a customer to a telecommunications network and a cloud environment to allow the customer to access multiple instances within the cloud through the connection. A customer may purchase multiple cloud resource instances from a public cloud environment and, utilizing the telecommunications network, connect to the multiple instances through a communication port or connection to the cloud environment. To utilize the single connection or port, communication packets intended for the cloud environment may be tagged with a VLAN tag that indicates to which cloud instance the packet is intended. The telecommunications network may route the packet to the intended cloud environment and configure one or more aspects of the cloud environment to analyze the attached VLAN tag to transmit the packet to the intended instance.

Abstract: A re-stacker assembly used to stack container closures that are output by a can liner system includes at least one offloading assembly having a receiving portion, at least one loading device coupled to the offloading assembly, and a motion control system for detecting motion of the container closures and, in response to detecting such motion, driving the loading device to stack, or re-stack, the container closures in a predetermined manner.

Abstract: There is provided a display device having a configuration suitable for increasing a screen size and reducing a thickness while ensuring adequate strength. The display device includes: a frame structure having a plurality of extrusion molded members coupled to one another; a display panel; and an intermediate member attached to the frame structure and holding the display panel.

Abstract: Systems and associated methods for recycling and performing other processes with consumer electronic devices are described herein. In various embodiments, the present technology includes systems and methods for identifying and evaluating a used or pre-owned consumer electronic device, such as a mobile phone, laptop, etc. to facilitate purchasing the device. In some embodiments, the present technology includes a counter-top evaluation terminal that evaluates a device and facilitates purchase and recycling of the device. Various other aspects of the present technology are described herein.

Abstract: Example embodiments of systems and methods for data transmission between a contactless card and a client device in support of a FIDO authentication are provided. In an embodiment, upon receipt of a challenge issued by a server in connection with a pending transaction, the contactless card may authorize the client device to utilize a FIDO private key to respond to the challenge. If the response to the challenge is successful, the FIDO authentication may proceed and the transaction may be completed.