Abstract: A method of processing audio is disclosed. The method includes capturing, via a recording device of an apparatus, an audio signal, wherein the audio signal includes background noise. The method also includes determining, at an uplink stage by a processor, whether the audio signal includes a voice segment of a speaker, and filtering, initially during the uplink stage, the audio signal to suppress or eliminate the background noise. The method further includes determining, by the processor, one or more parameters associated with the audio signal, and responsive to receiving the one or more parameters, filtering, at the downlink stage and based on the one or more parameters, an audio source associated with the audio signal that is played through a playback device of the apparatus during capturing of the audio signal, whereby the audio source includes the background noise of the audio signal captured by the recording device.

Abstract: Methods, systems, and computer programs encoded on a computer storage medium, for training and using machine learning models are disclosed. Methods include creating a model that represents relationships between user attributes, content exposures, and performance levels for a target action using organic exposure data specifying one or more organic exposures experienced by a particular user over a specified time prior to performance of a target action by the particular user and third party exposure data specifying third party exposures of a specified type of digital component to the particular user over the specified time period. Using the model, an incremental performance level attributable to each of the third party exposures at an action time when the target action was performed by the particular user is determined. Transmission criteria for at least some digital components to which the particular user was exposed are modified based on the incremental performance.

Abstract: Disclosed is a massive data-driven method for automatically locating a mine microseismic source, including: constructing a microseismic wave calibration data set by using a large-scale seismic data set containing seismic signals and non-seismic signals; constructing a pre-training calibration model based on a full convolution neural network through deep learning of a seismic wave calibration data set; using microseismic data of mine sites for transfer learning of an initial arrival time calibration model to construct an arrival time automatic calibration model suitable for mine microseismic signals; and automatically as well as accurately locating mine microseismic events based on an isokinetic homogeneous isotropic velocity model by using an optimization algorithm to deduce arrival time errors and through repeated iteration and fine-tuning.

Abstract: A method for training a coal rock interface recognition model is performed by a cloud server. The method includes receiving a sample coal rock distribution and sample multi-modal data from an edge processor, acquiring a sample load state feature based on the sample vibration data and the sample noise data; acquiring a sample cutting feature of a coal rock interface based on the sample video data. The method further includes acquiring a sample load feature for drum cutting based on the sample current data and the sample pressure data; calling the coal rock interface recognition model, and performing a decision-level fusion based on the sample load state feature; and training the coal rock interface recognition model based on the sample predicted coal rock distribution and the sample coal rock distribution to obtain a target coal rock interface recognition model.

Abstract: Methods, systems, and computer programs encoded on a computer storage medium, for training and using machine learning models are disclosed. Methods include creating a model that represents relationships between user attributes, content exposures, and performance levels for a target action using organic exposure data specifying one or more organic exposures experienced by a particular user over a specified time prior to performance of a target action by the particular user and third party exposure data specifying third party exposures of a specified type of digital component to the particular user over the specified time period. Using the model, an incremental performance level attributable to each of the third party exposures at an action time when the target action was performed by the particular user is determined. Transmission criteria for at least some digital components to which the particular user was exposed are modified based on the incremental performance.

Abstract: An integrated circuit device includes a storage circuit for storing a canceled value of a key identifier that is associated with a debug key used to sign debugging firmware. A security controller circuit is configurable to reactivate the key identifier to permit the debugging firmware to be loaded to the integrated circuit device to perform debugging functions by changing the canceled value of the key identifier stored in the first storage circuit to a reactivated value.

Abstract: A method of audio transmission of a voice segment of a speaker is disclosed. The method includes obtaining, via an encoder, the voice segment of the speaker. The method also includes obtaining, from one or more servers in communication with the encoder, a reference sample associated with the voice segment of the speaker. The method further includes establishing a timestamp associated with the voice segment of the speaker relative to the reference sample and encoding, via the encoder, an audio packet that includes the voice segment of the speaker and the timestamp associated with the voice segment of the speaker. The method also includes transmitting, in real-time, the audio packet to a decoder, wherein the voice segment of the speaker is to be combined with the reference sample based on the timestamp associated with the voice segment of the speaker during decoding.

Abstract: The present disclosure provides a manufacturing method of a semiconductor structure and a semiconductor structure. The manufacturing method includes: providing a substrate, where the substrate includes a complete die region and an incomplete die region; forming a stack on the substrate, where the stack includes sacrificial layers and supporting layers; forming a first photoresist layer on the stack; exposing the first photoresist layer, and developing to remove the first photoresist layer on the incomplete die region; and etching the stack by using the first photoresist layer on the complete die region as a mask.

Abstract: Embodiments provide a method for fabricating an array structure of a columnar capacitor and a semiconductor structure, relating to the field of semiconductor manufacturing technology. In the method, before a mask layer is removed, a thickness of the mask layer in the peripheral region is first adjusted to be equal to a thickness of the mask layer in the array region, thereby avoiding damage to a top support layer caused by different thicknesses of the mask layer. Moreover, in the method, a thickness of the top support layer is increased by means of a supplementary support layer, to increase support strength of the top support layer, thereby further preventing occurrence of tilt of the columnar capacitor due to insufficient support strength of the top support layer.

Abstract: This application relates to the field of internet technologies, and in particular, to a document display control method and apparatus, an electronic device, a storage medium, and a program product. The method includes: displaying, in a document region, one or more content blocks of a target document in a first page mode; and switching the document processing interface from the first page mode to a second page mode, and displaying the one or more content blocks of the target document in the second page mode, one of the first page mode and the second page mode being configured to present the one or more content blocks in a first preset order, the other being configured to present the one or more content blocks of the target document based on one or more objects, each of the one or more objects comprising at least one content block.

Abstract: Aspects of the disclosure are directed to a data transfer method that can include obtaining a first data packet reported by a first node device; receiving a second data packet forwarded by a signal amplifier, wherein the second data packet is reported by a second node device to the signal amplifier; processing the first data packet and the second data packet to obtain target data information; transferring the target data information to a server.

Abstract: A device includes a memory configured to store a collection of sets of weights, each of the sets of weights representing a respective media segment. The device also includes one or more processors configured to generate data representing the detected first input speech segment and to pass the data representing the detected first input speech segment into a collection of memory units. Each memory unit of the collection of memory units includes a set of weights from the collection of sets of weights. The one or more processors are also configured to generate a first estimate of an associated media segment that represents the detected first input speech segment. The associated media segment corresponds to a first memory unit in the collection of memory units.

Abstract: The present disclosure relates to the technical field of semiconductors, and provides a semiconductor structure and a manufacturing method thereof. The semiconductor structure includes: a substrate; a bit line located on the substrate; and a support layer located on the substrate, wherein the support layer includes a first support segment and a second support segment, the first support segment and the second support segment are both connected to the bit line, and the bit line is located between the first support segment and the second support segment.

Abstract: The present disclosure provides a manufacturing method of a semiconductor structure, including: an insulating layer includes a first dielectric layer and a second dielectric layer, a protective layer covers an upper surface of the second dielectric layer and a bottom and sidewalls of the first trench; removing part of the protective layer to expose at least part of a surface of the second dielectric layer; removing the second dielectric layer by a first wet etching process, the first wet etching process has a first etch selectivity of a material of the second dielectric layer to that of the first dielectric layer; and removing the protective layer by a second wet etching process, the second wet etching process has a second etch selectivity of a material of the protective layer to that of the first dielectric layer, and the second etch selectivity is greater than the first etch selectivity.

Abstract: A method of manufacturing a semiconductor structure and a semiconductor structure are disclosed. The method of manufacturing a semiconductor structure includes: providing a substrate, and forming a first sacrificial layer on the substrate, where the first sacrificial layer includes a first sacrificial dielectric layer and a second sacrificial dielectric layer; patterning the first sacrificial layer, and forming first intermediate pattern structures that are arranged at intervals, where a first gap is provided between two adjacent first intermediate pattern structures; forming a first spacer pad layer in the first gap, where the first spacer pad layer covers sidewalls of each of the two adjacent first intermediate pattern structures and a bottom of the first gap; removing the first spacer pad layer at the bottom of the first gap, and the second sacrificial dielectric layer; and removing the first sacrificial dielectric layer, to form first pattern structures.

Abstract: An embodiment of the application provides a method for forming a semiconductor structure. The semiconductor structure includes a first region and a second region. The method includes the following steps: providing a base, an insulating layer, and a mask layer that are stacked in sequence, where the first region has at least one trench penetrating the mask layer and the insulating layer, and the mask layer has an upper surface in the second region higher than that in the first region; forming a first protection layer, where an upper surface and a sidewall of the mask layer in the first region are covered with the first protection layer; after the first protection layer is formed, removing the mask layer in the second region; subsequent to removal of the mask layer in the second region, removing the first protection layer; and removing the mask layer in the first region.

Abstract: A device includes one or more processors configured to input one or more segments of an input media stream into a feature extractor. The one or more processors are further configured to pass an output of the feature extractor into an utterance classifier to produce at least one representation of at least one utterance class of a plurality of utterance classes. The one or more processors are further configured to pass the output of the feature extractor and the at least one representation into a segment matcher to produce a media output segment identifier.

Abstract: A method for preparing a semiconductor structure, and a semiconductor structure are provided. In a prepared first pattern structure, a thickness of a first insulating layer is equal to a thickness of a second insulating layer, and a thickness of a third insulating layer is equal to a thickness of a fourth insulating layer.

Abstract: A method for training a coal rock interface recognition model is performed by a cloud server. The method includes receiving a sample coal rock distribution and sample multi-modal data from an edge processor, acquiring a sample load state feature based on the sample vibration data and the sample noise data; acquiring a sample cutting feature of a coal rock interface based on the sample video data.

Abstract: A method for manufacturing a mask structure includes: patterning a sacrificial layer and a second dielectric layer, so as to form pattern structures each including a first pattern and a second pattern, and a width of a lower portion of the pattern structures is less than a width of a upper portion of the pattern structures; forming an initial mask pattern on sidewalls of each of the plurality of pattern structures; filling a first filling layer between adjacent initial mask patterns located on the sidewalls of different pattern structures; removing the second patterns and the initial mask pattern located on sidewalls of each of the plurality of second patterns; removing the first filling layer and the first patterns, so as to form first mask patterns; and forming second mask patterns on the first mask patterns.