Abstract: Systems and methods for dynamically altering shared content are provided. A set of media items to be shared with users of a content sharing service is identified. The set of media items includes a first media item associated with a first media item type and multiple second media items each associated with a second media item type. Each of the second media items correspond to at least one of a geographic location or a content preference associated with client devices connected to the content sharing service. In response to receiving a request for content from a first client device of the client devices connected to the content sharing service, a second media item is identified among the second media items. The identified second media item corresponds to a geographic location or a content preference associated with the first client device.

Abstract: Systems and method for routing data packets in an interconnection network. The data packets transmitted across the interconnection network each include age data. Routers positioned throughout the interconnection network may control the flow of the data packets through the use of aging first-in, first-out (FIFO) queues and age-based arbiters. The age-based arbiters within the routers are configured to prioritize older data packets over newer data packets being pushed from the FIFO queues. Each data packet inserted into the FIFO queues may be updated such that the age data of the data packet is converted from an age to an injection time. When the data packet is read from the FIFO, the age data of the data packet is converted back to an age.

Abstract: A system that includes: an array of qubits, each qubit of the array of qubits comprising a first electrode corresponding to a first node and a second electrode corresponding to a second node, wherein, for a first qubit in the array of qubits, the first qubit is positioned relative to a second qubit in the array of qubits such that a charge present on the first qubit induces a same charge on each of the first node of the second qubit and the second node of the second qubit, such that coupling between the first qubit and the second qubit is reduced, and wherein none of the nodes share a common ground is disclosed.

Abstract: Secure methods, systems, and media for generating and verifying user credentials are provided. In some embodiments, the method comprises: receiving, from a user device, a request for access to a service that requires valid user credentials; determining an aspect of the user credentials that is to be satisfied to grant access to the requested service; transmitting, to the user device, a request for information related to the aspect of the user credential; receiving, from the user device, information related to the aspect of the user credential, wherein the information has been signed using a key associated with the user device; verifying the key used to sign the information by the user device; in response to verifying the key used to sign the information, determining whether the aspect of the user credential has been satisfied based on the received information; and, in response to determining that the aspect of the user credential has been satisfied, granting access to the service.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for detecting objects in images. One of the methods includes obtaining an input image; processing the input image to generate predicted part feature data, the predicted part feature data comprising, for each of a plurality of possible object parts: a part presence probability representing a likelihood that the possible object part is depicted in the input image, a predicted pose of the possible object part in the input image given that the possible object part is depicted in the input image, and an object part feature vector characterizing the depiction of the possible object part given that the possible object part is depicted in the input image; and processing the predicted part feature data for the plurality of possible object parts to generate an object detection output that identifies one or more objects depicted in the input image.

Abstract: In one general aspect, a method can include capturing first features associated with a real-world physical area as a model and associating an AR object with a fixed location within the model. The method can include capturing second features associated with a real-world location corresponding with a portion of the real-world physical area. The method can include associating the captured second features with a location in the model, corresponding with the real-world location, as an AR anchor where the AR object is associated with the AR anchor.

Abstract: A virtual session manager of an electronic device maintains a web session for a user across multiple electronic devices. The virtual session manager receives an authentication request from a first electronic device that is in a communication range of the device. The virtual session manager transmits the authentication request to an endpoint device with a grant token without providing the first electronic device with any access to the grant token. The virtual session manager will receive, from the endpoint device, a first access token in response to the first authentication request. The virtual session manager will transmit the first access token to the first electronic device so that the first electronic device can establish a virtual session with the first web resource.

Abstract: Methods, systems, and apparatus, including a method for determining network measurements. In some aspects, a method includes receiving, by a first aggregation server and from each of multiple client devices, encrypted impression data. A second aggregation server receives, from each of at least a portion of the multiple client devices, encrypted conversion data. The first aggregation server and the second aggregation server perform a multi-party computation process to generate chronological sequences of encrypted impression data and encrypted conversion data and to decrypt the encrypted impression data and the encrypted conversion data.

Abstract: A method includes receiving text input data for conversion into synthesized speech and determining, using a hotword-aware model trained to detect a presence of a hotword assigned to a user device, whether a pronunciation of the text input data includes the hotword. The hotword is configured to initiate a wake-up process on the user device for processing the hotword and/or one or more other terms following the hotword in the audio input data. When the pronunciation of the text input data includes the hotword, the method also includes generating an audio output signal from the text input data and providing the audio output signal to an audio output device to output the audio output signal. The audio output signal when captured by an audio capture device of the user device, configured to prevent initiation of the wake-up process on the user device.

Abstract: This application is directed to a speaker assembly in which a speaker is mounted in an enclosure structure. The enclosure structure exposes a speaker opening of the speaker and provides a sealed enclosure for a rear portion of the speaker, and further includes an electrically conductive portion. One or more electronic components are coupled to the electrically conductive portion of the enclosure structure (which is grounded in some implementations). The electrically conductive portion of the enclosure structure is configured to provide electromagnetic shielding for the electronic components and forms part of the sealed enclosure of the speaker. In some implementations, the electrically conductive portion of the enclosure structure is thermally coupled to the electronic components and acts as a heat sink that is configured to absorb heat generated by the electronic components and dissipate the generated heat away from the electronic components.

Abstract: Various arrangements for performing the playback of media content are provided. A computing device can cause a media content item to be transmitted to a display device for presentation. An event that is associated with the presentation of the media content item can be detected. In response to the event, a mirroring session can be resumed on the computing device.

Abstract: A direct speech-to-speech translation (S2ST) model includes an encoder configured to receive an input speech representation that to an utterance spoken by a source speaker in a first language and encode the input speech representation into a hidden feature representation. The S2ST model also includes an attention module configured to generate a context vector that attends to the hidden representation encoded by the encoder. The S2ST model also includes a decoder configured to receive the context vector generated by the attention module and predict a phoneme representation that corresponds to a translation of the utterance in a second different language. The S2ST model also includes a synthesizer configured to receive the context vector and the phoneme representation and generate a translated synthesized speech representation that corresponds to a translation of the utterance spoken in the different second language.

Abstract: A system for passive alignment of fibers to an interface of a photonic integrated circuit (PIC) includes an input frame, an actuator, and an output frame. The actuator arranged to apply force along a force axis to the input frame. The output frame including a tip for picking up a plate and transferring the force thereto, the output frame being connected to the input frame such that the output frame may tilt relative to the input frame and the output frame is elastically biased relative to the input frame into a position wherein the tip is aligned on the force axis.

Abstract: Described techniques enable displaying a plurality of positions arranged around a central location on a display, the plurality of positions including at least one empty position and filled positions displaying a first subset of list elements, the filled positions including an adjacent position that is adjacent to the at least one empty position. Movement of a gaze point may be tracked, relative to the filled positions and around the central location towards the at least one empty position. The at least one empty position may be filled with a list element of the list elements, and the adjacent position may be emptied to thereby display a second subset of the list elements.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for coding speech using neural networks. One of the methods includes obtaining a bitstream of parametric coder parameters characterizing spoken speech; generating, from the parametric coder parameters, a conditioning sequence; generating a reconstruction of the spoken speech that includes a respective speech sample at each of a plurality of decoder time steps, comprising, at each decoder time step: processing a current reconstruction sequence using an auto-regressive generative neural network, wherein the auto-regressive generative neural network is configured to process the current reconstruction to compute a score distribution over possible speech sample values, and wherein the processing comprises conditioning the auto-regressive generative neural network on at least a portion of the conditioning sequence; and sampling a speech sample from the possible speech sample values.

Abstract: A method includes providing, via a user interface (UI) of a first application on a first screen device, a UI element prompting a user to connect the first screen device with a second screen device that is presenting a video to the user. The method further includes receiving, via the UI element, an indication of a user request to connect the first screen device with the second screen device. In response to receiving the indication of the user request to connect the first screen device with the second screen device, the method further includes causing the first screen device to be paired with the second screen device. The method further includes providing, for presentation in the UI of the first application of the first screen device, on or more comments provided by one or more other users for the video that is concurrently presented on the second screen device.

Abstract: A computer-implemented method that includes receiving, by a processing unit, an instruction that specifies data values for performing a tensor computation. In response to receiving the instruction, the method may include, performing, by the processing unit, the tensor computation by executing a loop nest comprising a plurality of loops, wherein a structure of the loop nest is defined based on one or more of the data values of the instruction. The tensor computation can be at least a portion of a computation of a neural network layer. The data values specified by the instruction may comprise a value that specifies a type of the neural network layer, and the structure of the loop nest can be defined at least in part by the type of the neural network layer.

Abstract: A recurrent neural network-transducer (RNN-T) model improves speech recognition by processing sequential non-blank symbols at each time step after an initial one. The model's prediction network receives a sequence of symbols from a final Softmax layer and employs a shared embedding matrix to create and map embeddings to each symbol, associating them with unique position vectors. These embeddings are weighted according to their similarity to their matching position vector. Subsequently, a joint network of the RNN-T model uses these weighted embeddings to output a probability distribution for potential speech recognition hypotheses at each time step, enabling more accurate transcriptions of spoken language.

Abstract: Systems and methods for validation of modeling and simulation systems that provide for the virtual fitting of wearable devices, such as glasses, by a user. Three-dimensional modeling and simulation of test subject both with and without fitting frames corresponding to a wearable device may be captured to validate the modeling and simulation modules and associated algorithms and machine learning modules used to simulate the fit of the wearable device on a user. Validation in this manner may provide for increased accuracy/realism of the modeling and simulation systems.