Abstract: Embodiments of the present disclosure provide a method, apparatus, device and storage medium for data processing. If a kernel module sends a data calling request to a userspace process, first modification time information and second modification time information of data corresponding to the data calling request are obtained by the userspace process. The first and second modification time information are used to indicate modification time information of the data in the kernel module and in a file service end, respectively. The first and second modification time information are compared by the userspace process. If the first and second modification time information are inconsistent, a verification invalidation result is returned to the kernel module. The data in the kernel module is invalidated by the kernel module according to the verification invalidation result. The data in the file service is synchronized by the userspace process to the kernel module.

Abstract: Embodiments of the present disclosure provide a method, apparatus, device and storage medium for data processing. If a kernel module sends a data calling request to a userspace process, first modification time information and second modification time information of data corresponding to the data calling request are obtained by the userspace process. The first and second modification time information are used to indicate modification time information of the data in the kernel module and in a file service end, respectively. The first and second modification time information are compared by the userspace process. If the first and second modification time information are inconsistent, a verification invalidation result is returned to the kernel module. The data in the kernel module is invalidated by the kernel module according to the verification invalidation result. The data in the file service is synchronized by the userspace process to the kernel module.

Abstract: Embodiments of the present disclosure provide a method, an apparatus, a device and a storage medium for data request processing. The method is applied at a filesystem in userspace which includes a kernel module and a userspace process. The method includes: obtaining a data request list in the kernel module, where the data request list includes a plurality of data requests to be processed and virtual address information corresponding to respective data requests of the plurality of data requests; querying, for each of the data requests, physical address information corresponding to the virtual address information, according to the virtual address information corresponding to the data request; querying target cached data corresponding to the data request, from a plurality of cached data stored in the kernel module; accessing the target cached data, and processing the data request.

Abstract: The present invention provides a DNA-targeting RNA comprising a single-guide RNA (sgRNA) and a ribonucleotide sequence rich in adenine ribonucleotide, and its use in gene editing, and a method for improving the efficiency of sgRNA-mediated gene editing, comprising a step of adding a ribonucleotide sequence rich in adenine ribonucleotide at 3? end of the sgRNA.

Abstract: Disclosed is a boronic acid derivative. Provided are a compound of formula (I) or a pharmaceutically acceptable salt, solvate, polymorph or isomer thereof, a pharmaceutical composition containing the same, and the use thereof in the treatment of lmp7-related diseases.

Abstract: Systems and methods are provided for generating a lightweight, high-quality streaming text-to-speech (TTS) system. For example, some disclosed systems are configured to obtain a first model comprising one or more layers of a convolutional neural network. Each layer of the convolutional neural network is configured to generate a new output from a previous input. The systems also obtain a second model comprising a recurrent neural network. Subsequent to obtaining the first model and the second model, the systems are configured to compile the one or more layers of the convolutional neural network and the recurrent neural network in a parallel architecture to generate a machine learning module such that each model of the machine learning module is configured to receive input simultaneously.

Abstract: Automatic generation of intelligent content is created using a system of computers including a user device and a cloud-based component that processes the user information. The system performs a process that includes receiving an input document and parsing the input document to generate inputs for a natural language generation model using a text analysis model. The natural language generation model generates one or more candidate presentation scripts based on the inputs. A presentation script is selected from the candidate presentation scripts and displayed. A text-to-speech model may be used to generate a synthesized audio presentation of the presentation script. A final presentation may be generated that includes a visual display of the input document and the corresponding audio presentation in sync with the visual display.

Abstract: According to implementations of the subject matter described herein, a solution is proposed for text to speech. In this solution, an initial phoneme sequence corresponding to text is generated, the initial phoneme sequence comprising feature representations of a plurality of phonemes. A first phoneme sequence is generated by inserting a feature representation of an additional phoneme into the initial phoneme sequence, the additional phoneme being related to a characteristic of spontaneous speech. The duration of a phoneme among the plurality of phonemes and the additional phoneme is determined by using an expert model corresponding to the phoneme, and a second phoneme sequence is generated based on the first phoneme sequence. Spontaneous-style speech corresponding to the text is determined based on the second phoneme sequence. In this way, spontaneous-style speech with more varying rhythms can be generated based on spontaneous-style additional phonemes and multiple expert models.

Abstract: Automatic generation of intelligent content is created using a system of computers including a user device and a cloud-based component that processes the user information. The system performs a process that includes receiving an input document and parsing the input document to generate inputs for a natural language generation model using a text analysis model. The natural language generation model generates one or more candidate presentation scripts based on the inputs. A presentation script is selected from the candidate presentation scripts and displayed. A text-to-speech model may be used to generate a synthesized audio presentation of the presentation script. A final presentation may be generated that includes a visual display of the input document and the corresponding audio presentation in sync with the visual display.

Abstract: An improved, efficient method for mapping world points from an environment (e.g., points generated by a LIDAR sensor of an autonomous vehicle) to locations (e.g., pixels) within rolling-shutter images taken of the environment is provided. This improved method allows for accurate localization of the world point in a rolling-shutter image via an iterative process that converges in very few iterations. The method poses the localization process as an iterative process for determining the time, within the rolling-shutter exposure period of the image, at which the world point was imaged by the camera. The method reduces the number of times the world point is projected into the normalized space of the camera image, often converging in three or fewer iterations.

Abstract: Systems and methods for extrinsic calibration of vehicle-mounted sensors are provided. One example method involves obtaining first sensor data collected by a first sensor and a second sensor while a vehicle is aligned in a first yaw direction. The method also involves obtaining second sensor data collected by the first sensor and the second sensor while the vehicle is aligned in a second yaw direction. The method also involves determining, based on the first sensor data and the second sensor data, (i) first pitch and roll misalignments of the first sensor relative to the vehicle and (ii) second pitch and roll misalignments of the second sensor relative to the first sensor. The method also involves determining third pitch and roll misalignments of the second sensor relative to the vehicle based on (i) the first pitch and roll misalignments and (ii) the second pitch and roll misalignments.

Abstract: A method and system for enhancing pronunciation during a speech, the method including receiving audio data, the audio data including a speech, performing at least one of acoustic scoring and language scoring on the speech, determining a pronunciation score of one or more words of the speech based on the acoustic scoring and the language scoring, determining that the pronunciation score for the word does not satisfy a threshold score, responsive to determining that the pronunciation score does satisfy the threshold score, identifying the word as mispronounced, and responsive to identifying the word as mispronounced, outputting the word and the pronunciation score thereof.

Abstract: An electronic device tracks its motion in an environment while building a three-dimensional visual representation of the environment that is used to correct drift in the tracked motion. A motion tracking module estimates poses of the electronic device based on feature descriptors corresponding to the visual appearance of spatial features of objects in the environment. A mapping module builds a three-dimensional visual representation of the environment based on a stored plurality of maps, and feature descriptors and estimated device poses received from the motion tracking module. The mapping module provides the three-dimensional visual representation of the environment to a localization module, which identifies correspondences between stored and observed feature descriptors. The localization module performs a loop closure by minimizing the discrepancies between matching feature descriptors to compute a localized pose.

Abstract: Implementations of the subject matter described herein provide a solution for speech synthesis and speech recognition. In this solution, a Text to Speech (TTS) model and an Automatic Speech Recognition (ASR) model supporting at least one language are obtained. The TTS model and the ASR model are adjusted, based on a first set of paired data in a target language, to support the target language. The TTS model is optimized based on the first set of paired data and a first set of synthesized paired data in the target language generated by the ASR model while the ASR model is optimized based on the first set of paired data and a second set of synthesized paired data in the target language generated by the TTS model. As such, the solution can provide TTS and ASR models with high accuracy for languages lacking training data by using less training data.

Abstract: An electronic device tracks its motion in an environment while building a three-dimensional visual representation of the environment that is used to correct drift in the tracked motion. A motion tracking module estimates poses of the electronic device based on feature descriptors corresponding to the visual appearance of spatial features of objects in the environment. A mapping module builds a three-dimensional visual representation of the environment based on a stored plurality of maps, and feature descriptors and estimated device poses received from the motion tracking module. The mapping module provides the three-dimensional visual representation of the environment to a localization module, which identifies correspondences between stored and observed feature descriptors. The localization module performs a loop closure by minimizing the discrepancies between matching feature descriptors to compute a localized pose.

Abstract: A radar system and a terminal device are provided. The radar system includes a controller and at least two radar modules directly or indirectly connected to the controller. The at least two radar modules include a first radar module and a second radar module, and the first radar module and the second radar module implement time division multiplexing of the controller in a digital domain. Compared with an existing radar system, the radar system in this application can provide more transmit channels, more receive channels, and a larger antenna array size when the two radar systems include a same quantity of controllers.

Abstract: The described aspects and implementations enable efficient calibration of a sensing system of an autonomous vehicle (AV). In one implementation, disclosed is a method and a system to perform the method, the system including the sensing system configured to collect sensing data and a data processing system, operatively coupled to the sensing system. The data processing system is configured to identify reference point(s) in an environment of the AV, determine multiple estimated locations of the reference point(s), and adjust parameters of the sensing system based on a loss function representative of differences of the estimated locations.

Abstract: Systems are configured for generating spectrogram data characterized by a voice timbre of a target speaker and a prosody style of source speaker by converting a waveform of source speaker data to phonetic posterior gram (PPG) data, extracting additional prosody features from the source speaker data, and generating a spectrogram based on the PPG data and the extracted prosody features. The systems are configured to utilize/train a machine learning model for generating spectrogram data and for training a neural text-to-speech model with the generated spectrogram data.

Abstract: Novel solutions for speech recognition provide contextual spelling correction (CSC) for automatic speech recognition (ASR). Disclosed examples include receiving an audio stream; performing an ASR process on the audio stream to produce an ASR hypothesis; receiving a context list; and, based on at least the ASR hypothesis and the context list, performing spelling correction to produce an output text sequence. A contextual spelling correction (CSC) model is used on top of an ASR model, precluding the need for changing the original ASR model. This permits run-time user customization based on contextual data, even for large-size context lists. Some examples include filtering ASR hypotheses for the audio stream and, based on at least the ASR hypotheses filtering, determining whether to trigger spelling correction for the ASR hypothesis. Some examples include generating text to speech (TTS) audio using preprocessed transcriptions with context phrases to train the CSC model.

Abstract: Automatic generation of intelligent content is created using a system of computers including a user device and a cloud-based component that processes the user information. The system performs a process that includes receiving an input document and parsing the input document to generate inputs for a natural language generation model using a text analysis model. The natural language generation model generates one or more candidate presentation scripts based on the inputs. A presentation script is selected from the candidate presentation scripts and displayed. A text-to-speech model may be used to generate a synthesized audio presentation of the presentation script. A final presentation may be generated that includes a visual display of the input document and the corresponding audio presentation in sync with the visual display.