Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for optimizing the execution of the operations of a neural network. One of the methods includes obtaining data representing a graph characterizing a plurality of operations of a neural network, wherein each node of the graph characterizes an operation of the neural network and each edge of the graph characterizes data dependency between the operations; processing the data representing the graph using a graph embedding neural network to generate an embedding of the graph; and processing the embedding of the graph using a policy neural network to generate a task output, wherein the task output comprises, for each of the plurality of operations of the neural network, a respective decision for a particular optimization task.

Abstract: A processing service of a provider network may protect media content from being tampered with when it is transmitted from the provider network/transcoder to untrusted networks (e.g., third-party networks/CDNs) and to a media player. The processing service (e.g., the transcoder) generates a public and a private key. The service uses the private key to digitally sign content portions (e.g., video frames) before distribution to untrusted CDNs. The provider network creates a manifest that includes the public key. To play the media content, the media player obtains a manifest that includes the public key (via a secure/trusted connection with the provider network). The media player may then obtain the media content from an untrusted edge server/CDN and validate it using the public key that was separately obtained from the manifest (to verify the content was not tampered with).

Abstract: A processing service of a provider network may protect media content from being tampered with when it is transmitted from the provider network/transcoder to untrusted networks (e.g., third-party networks/CDNs) and to a media player. The processing service (e.g., the transcoder) generates a public and a private key. The service uses the private key to digitally sign content portions (e.g., video frames) before distribution to untrusted CDNs. The provider network creates a manifest that includes the public key. To play the media content, the media player obtains a manifest that includes the public key (via a secure/trusted connection with the provider network). The media player may then obtain the media content from an untrusted edge server/CDN and validate it using the public key that was separately obtained from the manifest (to verify the content was not tampered with).

Abstract: A processing service of a provider network may protect media content from being tampered with when it is transmitted from the provider network/transcoder to untrusted networks (e.g., third-party networks/CDNs) and to a media player. The processing service (e.g., the transcoder) generates a public and a private key. The service uses the private key to digitally sign content portions (e.g., video frames) before distribution to untrusted CDNs. The provider network creates a manifest that includes the public key. To play the media content, the media player obtains a manifest that includes the public key (via a secure/trusted connection with the provider network). The media player may then obtain the media content from an untrusted edge server/CDN and validate it using the public key that was separately obtained from the manifest (to verify the content was not tampered with).

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for optimizing the execution of the operations of a neural network. One of the methods includes obtaining data representing a graph characterizing a plurality of operations of a neural network, wherein each node of the graph characterizes an operation of the neural network and each edge of the graph characterizes data dependency between the operations; processing the data representing the graph using a graph embedding neural network to generate an embedding of the graph; and processing the embedding of the graph using a policy neural network to generate a task output, wherein the task output comprises, for each of the plurality of operations of the neural network, a respective decision for a particular optimization task.

Abstract: A processing service of a provider network may protect media content from being tampered with when it is transmitted from the provider network/transcoder to untrusted networks (e.g., third-party networks/CDNs) and to a media player. The processing service (e.g., the transcoder) generates a public and a private key. The service uses the private key to digitally sign content portions (e.g., video frames) before distribution to untrusted CDNs. The provider network creates a manifest that includes the public key. To play the media content, the media player obtains a manifest that includes the public key (via a secure/trusted connection with the provider network). The media player may then obtain the media content from an untrusted edge server/CDN and validate it using the public key that was separately obtained from the manifest (to verify the content was not tampered with).

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for compiler optimizations using a compiler optimization network. One of the methods includes receiving an input program, wherein the input program defines a graph of operation modules, wherein each node in the graph is a respective operation module, and each edge between nodes in the graph represents one operation module receiving the output generated by another operation module. The input program is processed by a compiler optimization network comprising a graph-embedding network that is configured to encode operation features and operation dependencies of the operation modules of the input program into a graph embedding representation and a policy network that is configured to generate an optimization action for each of one or more nodes encoded in the graph embedding representation.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for optimizing the execution of the operations of a neural network. One of the methods includes obtaining data representing a graph characterizing a plurality of operations of a neural network, wherein each node of the graph characterizes an operation of the neural network and each edge of the graph characterizes data dependency between the operations; processing the data representing the graph using a graph embedding neural network to generate an embedding of the graph; and processing the embedding of the graph using a policy neural network to generate a task output, wherein the task output comprises, for each of the plurality of operations of the neural network, a respective decision for a particular optimization task.

Abstract: A vehicle airbag system includes an airbag module, a seat detector, and an electronic controller. The airbag module has an airbag that is configured to be movably mounted to a vehicle roof structure. The seat detector detects a position of a vehicle seat. The electronic controller is in communication with the airbag module and the seat detector. The controller controls the airbag module to inflate the airbag in a predetermined direction based on detected position of the vehicle seat by the seat detector.

Abstract: A processing service of a provider network may protect media content from being tampered with when it is transmitted from the provider network/transcoder to untrusted networks (e.g., third-party networks/CDNs) and to a media player. The processing service (e.g., the transcoder) generates a public and a private key. The service uses the private key to digitally sign content portions (e.g., video frames) before distribution to untrusted CDNs. The provider network creates a manifest that includes the public key. To play the media content, the media player obtains a manifest that includes the public key (via a secure/trusted connection with the provider network). The media player may then obtain the media content from an untrusted edge server/CDN and validate it using the public key that was separately obtained from the manifest (to verify the content was not tampered with).

Abstract: Systems and methods for conducting a payment transaction at a vending machine using a mobile device are disclosed herein. A vending machine includes a Bluetooth Low Energy (BLE) module that is used to connect to a user's mobile device. The mobile device's web browser accesses a web page that establishes a BLE connection between the vending machine and mobile device and also facilitates a payment transaction with an online payment service provider without transmitting any payment information to the vending machine.

Abstract: This disclosure provides a multi-layered printed circuit board (PC) that has signal array region. The signal array region has a width and circumscribes a power core region and has signal vias connected to respective signal ball pads, and ground vias connected to respective ground ball pads within the signal array region that have an associated ball pad pitch. The PCB also has an inner current power layer. The signal and ground vias are arranged on the component layer in a pattern and extend into the inner current layer. The pattern forms current power paths across the width of the signal array region, such that the current power paths have a width that is at least about 50% as wide as the ball pad pitch.

Abstract: Embodiments herein provide systems and methods tracking the weight of an individual situated on a furnishing such as a chair or bed. One method generally includes receiving weight data at a weight sensor that is situated to recognize a change in weight on the furnishing, determining that a mobile device of the individual is proximately located to the weight sensor, and confirming by the mobile device that the change in weight is related to the individual.

Abstract: Embodiments herein provide systems and methods tracking the weight of an individual situated on a furnishing such as a chair or bed. One method generally includes receiving weight data at a weight sensor that is situated to recognize a change in weight on the furnishing, confirming that a fitness device is proximately located to the weight sensor, and transmitting the received weight data to a hosted service.

Abstract: A battery material includes: a conductive substrate; and a silicon material layer disposed on at least one side of the conductive substrate and including silicon fibers, at least some of which aggregate to form silicon fibrous spheres. A preparation method for the battery material includes: (a) providing a conductive substrate; (b) providing a silicon-based material, followed by depositing on a surface of the silicon-based material a film comprising amorphous silicon or crystalline silicon to form a silicon source material; (c) placing the silicon source material of step (b) on at least a side of the conductive substrate of step (a), wherein the distance between the conductive substrate and the silicon source material is less than 100 mm; and (d) heating the conductive substrate and silicon source material thus arranged in step (c) to allow the silicon source material to release amorphous silicon or crystalline silicon.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for optimizing the execution of the operations of a neural network. One of the methods includes obtaining data representing a graph characterizing a plurality of operations of a neural network, wherein each node of the graph characterizes an operation of the neural network and each edge of the graph characterizes data dependency between the operations; processing the data representing the graph using a graph embedding neural network to generate an embedding of the graph; and processing the embedding of the graph using a policy neural network to generate a task output, wherein the task output comprises, for each of the plurality of operations of the neural network, a respective decision for a particular optimization task.

Abstract: This disclosure provides a multi-layered printed circuit board (PC) that has signal array region. The signal array region has a width and circumscribes a power core region and has signal vias connected to respective signal ball pads, and ground vias connected to respective ground ball pads within the signal array region that have an associated ball pad pitch. The PCB also has an inner current power layer. The signal and ground vias are arranged on the component layer in a pattern and extend into the inner current layer. The pattern forms current power paths across the width of the signal array region, such that the current power paths have a width that is at least about 50% as wide as the ball pad pitch.

Abstract: A solar cell includes a substrate, an emitter layer, a first thin oxide layer, a first doped silicon layer, a second thin oxide layer, and a second doped silicon layer. The substrate has a first side and a second side opposite to each other. The emitter layer is disposed on the first side of the substrate and includes a first conductivity type of dopant. The first thin oxide layer is disposed on the emitter layer. The first doped silicon layer is disposed on the first thin oxide layer and includes the first conductivity type of dopant. The second thin oxide layer is disposed on the second side of the substrate. The second doped silicon layer is disposed on the second thin oxide layer and includes a second conductivity type of dopant. A method for manufacturing a solar cell is also provided herein.

Abstract: A vehicle airbag system includes an airbag module, a seat detector, and an electronic controller. The airbag module has an airbag that is configured to be movably mounted to a vehicle roof structure. The seat detector detects a position of a vehicle seat. The electronic controller is in communication with the airbag module and the seat detector. The controller controls the airbag module to inflate the airbag in a predetermined direction based on detected position of the vehicle seat by the seat detector.

Abstract: Embodiments herein provide systems and methods tracking the weight of an individual situated on a furnishing such as a chair or bed. One method generally includes receiving weight data at a weight sensor that is situated to recognize a change in weight on the furnishing, confirming that a fitness device is proximately located to the weight sensor, and transmitting the received weight data to a hosted service.