Abstract: Systems and methods are provided for automatically determining an intent of a user based on an intent model to attach a file to a document, prompting the user to confirm the intent using a predetermined character in an inline nudge, generating and displaying an inline menu with an interactive list of ranked files as a suggestion for attachment. The disclosed technology uses the intent for specifying a scope of the inline search. The intent model for attaching content maintained by third-party applications uses a combination of an embeddings model and an N-gram model with limited seed queries and determines the intent based on intent scores associated with respective third-party applications. The present disclosure ranks respective candidate content based on a degree of relevance to the intent. The user selects one or more content from the list for attaching to the document.

Abstract: Described herein are techniques, devices, and systems for selectively using a music-capable audio codec on-demand during a communication session. A user equipment (UE) may adaptively transition between using a first audio codec that provides a first audio bandwidth and a second audio codec (e.g., the EVS-FB codec) that provides a second audio bandwidth that is greater than the first audio bandwidth. The transition to the second audio codec may occur in response to determining that sound in the environment of the UE includes frequencies outside of a range of frequencies associated with a human voice, such as by determining that music is being played in the environment of the UE, which allows for selectively using a music-capable audio codec when it would be beneficial to do so.

Abstract: An air quality and particulate detection system comprising: an air quality and particulate detector comprising a particulate detection chamber in fluid communication via an orifice with an adjacent air quality member, the particulate detection chamber comprising a blower adjacent to a particle filter and a laser source, and the air quality member comprising a channel having substantially parallel sides for operably coupling a HEPA filter adjacent and substantially parallel to the orifice, and for operably coupling one or more ultraviolet light emitting diodes (LEDs) downstream from the HEPA filter, and an exhaust pipe for returning treated air to a protected area; one or more pipe runs fluidly connected to the particulate detection chamber, wherein each of the one or more pipe runs includes one or more sampling points configured to allow air particulate entry from the protected area, into at least one of the one or more pipe runs.

Abstract: An electronic apparatus includes a receiver comprising input receiving circuitry, a speaker, a display, and a processor configured to: based on a speech being received, determine a vibe keyword corresponding to a vibe-related word included in the speech among a plurality of specified vibe keywords, generate an image and a sound corresponding to the speech based on a graphical user interface (GUI) component and a sound component corresponding to the determined vibe keyword, and display the generated image on the display and output the generated sound through the speaker.

Abstract: Described herein are systems and methods for training machine learning models to generate three-dimensional (3D) motions based on light detection and ranging (LiDAR) point clouds. In various embodiments, a computing system can encode a machine learning model representing an object in a scene. The computing system can train the machine learning model using a dataset comprising synchronous LiDAR point clouds captured by monocular LiDAR sensors and ground-truth three-dimensional motions obtained from IMU devices. The machine learning model can be configured to generate a three-dimensional motion of the object based on an input of a plurality of point cloud frames captured by a monocular LiDAR sensor.

Abstract: In one embodiment, the disclosure provides a method on a drilling fluid. The method may include acquiring data from each of a particle-size analyzer, x-ray fluorescence instrument, a mud check, and optionally an x-ray diffraction instrument, to yield acquired data for each of a plurality of drilling mud samples, wherein: (a) the particle size analyzer measures a plurality of drilling mud samples; (b) the x-ray fluorescence instrument and the x-ray diffraction instrument measure a mineral analysis of the plurality of drilling mud samples; and (c) the mud check measures properties of the plurality of drilling mud samples. Further, the method may include plotting a particle size distribution curve for at least one of the plurality of drilling mud samples, and calculating an amount of abrasive formation material(s) within at least one of the plurality of drilling mud samples. Further still, the method may include adjusting the drilling fluid.

Abstract: Origin text content to be analyzed using natural language processing is received. A two-dimensional item sequence representation for at least a portion of the received origin text content is generated. Using one or more processors, one or more evaluation metrics are determined based on an analysis of the two-dimensional item sequence representation. A reduced version of the origin text content is automatically generated based on the one or more evaluation metrics to assist in satisfying a constraint of a natural language processing model. The reduced version of the origin text content is used as an input to the natural language processing model.

Abstract: The embodiments include systems and methods to determine torque and drag of a string. A method includes analyzing a plurality of continuous segments of a string deployed in a wellbore, and determining a deflection of the plurality of continuous segments from a node of the string and a tortuosity deflection of the wellbore. In response to a determination that deflection of the plurality of continuous segments from the node of the string is greater than the tortuosity deflection of the wellbore, the method includes applying a soft string model to determine a torque and a drag of the string. In response to a determination that deflection of the plurality of continuous segments from the node of the string is not greater than the tortuosity deflection of the wellbore, the method includes applying a stiff string model to determine a torque and a drag of the string.

Abstract: An information processing apparatus includes: a first acquisition unit that acquires odor data obtained by measuring the odor of an object; a second acquisition unit that acquires an acquisition condition for the odor data; and an identification unit that identifies the object from the odor data and the acquisition condition acquired by the first and second acquisition units on the basis of a learned model obtained by performing learning on the odor data of the object and the acquisition condition, and on the object corresponding to the odor data. The acquisition condition can be text data that indicates the category of the odor and is inputted by a user who has measured the odor.

Abstract: Systems and methods for the multifaceted analysis of a written work, such as an educational program admission essay, using machine learning, deep learning and natural language processing. Language, relevance, structure, and flows are evaluated for an overall impactful essay. Essay content is checked to evaluate whether the author has covered an essay's essential aspects. The essay is also analyzed for an effective structure for presenting details as per the essay type. The disclosure includes data preparation for the task, process of data tagging, feature engineering from the essay text, method for transfer learning and fine-tuning language model to adapt to the context of an essay. Finally, a process for building machine learning and deep learning models and technique for ensembling to use both models in combination is disclosed. The system may provide user-adapted feedback based on a persona created from the user profile.

Abstract: A network threat intelligence relational triple combined extraction method based on deep learning includes: (a) an entity and a relation are both extracted by using a combined extraction method, which solves the problem of a lack of interaction between entity and relation extraction tasks; (b) a problem of entity overlapping is solved by using a method based on a span; (c) a BERT large-scale pre-training model is used for vector representation of a text; and because the pre-training model includes contextual information learned from a large-scale corpus, semantic expression on a threat intelligence text by the model can be extremely enriched; and (d) various modal information, for example, time sequences, dependency relations, the spans, labels, and the like, is fused to enhance the interaction between multimodal information.

Abstract: Techniques for determining conditions of power lines in a power distribution system based on data collected by a plurality of sensor units deployed in the power distribution system. The techniques include obtaining data associated with measurements collected by at least two sensor units in the plurality of sensor units, and determining, by using at least one processor, at least one condition of at least one power line in the power distribution system by using the data obtained by the at least two sensor units.

Abstract: A system is provided for modifying how an output is presented via a multi-device synchronous configuration based on detecting a speech characteristic in the user input. For example, if the user whispers a request, then the system may temporarily modify how the responsive output is presented to the user via multiple devices. In one example, the system may lower the volume on all devices presented the output. In another example, the system may present the output via a single device rather than multiple devices. The system may also determine to operate in an alternate output mode based on certain non-audio data.

Abstract: Described is a way to process radar data for a radar sensor mounted on a vehicle to generate motion spectrum data for estimating ego-motion information of the vehicle. Data samples of each of a plurality of radar return signals received at each antenna element of the radar sensor are generated for each antenna element. Respective Doppler-processed data including a plurality of data values is calculated for each Doppler bin index. In generating a set of motion spectrum data, which comprises a plurality of data elements each calculated for a respective Doppler bin index and a respective spatial bin index, data values of the Doppler-processed data calculated for the Doppler bin index are selected to calculate a covariance matrix. A spectral estimation algorithm, which uses the covariance matrix, can determine a spatial spectrum value for each spatial bin index.

Abstract: Implementations can reduce the time required to obtain responses from an automated assistant through proactive caching, locally at a client device, of proactive assistant cache entries—and through on-device utilization of the proactive assistant cache entries. Different proactive cache entries can be provided to different client devices, and various implementations relate to technique(s) utilized in determining which proactive cache entries to provide to which client devices. In some of those implementations, in determining which proactive cache entries to provide (proactively or in response to a request) to a given client device, a remote system selects, from a superset of candidate proactive cache entries, a subset of the cache entries for providing to the given client device.

Abstract: A computer-implemented method of predicting an event horizon is disclosed. The method comprises maintaining condition data indicating a plurality of conditions occurring on one or more physical systems at a plurality of points in time. The method further comprises receiving an input feature vector representing a given condition occurring at a specific time during a specific period of time. The method also comprises generating, using a particular trained machine learning model of a plurality of trained machine learning models, a forecast value that indicates an amount of time from the specific time to an occurrence of a particular target condition on a particular physical system, the particular target condition being different from the given condition, each trained machine learning model corresponding to a distinct target condition. In addition, the method comprises causing, based on the forecast value, an action to be executed on the particular physical system.

Abstract: Techniques are described herein for detecting and/or enrolling (or commissioning) new “hot commands” that are useable to cause an automated assistant to perform responsive action(s) without having to be first explicitly invoked. In various implementations, an automated assistant may be transitioned from a limited listening state into a full speech recognition state in response to a trigger event. While in the full speech recognition state, the automated assistant may receive and perform speech recognition processing on a spoken command from a user to generate a textual command. The textual command may be determined to satisfy a frequency threshold in a corpus of textual commands. Consequently, data indicative of the textual command may be enrolled as a hot command. Subsequent utterance of another textual command that is semantically consistent with the textual command may trigger performance of a responsive action by the automated assistant, without requiring explicit invocation.

Abstract: The current document is directed to a methods and systems that use observational data collected by various devices and sensors to generate electronic-data representations of human conversations. The implementations of these methods and systems, disclosed in the current document, provide a highly extensible and generic platform for converting observational data into affect-annotated-timeline outputs that provide both a textual transcription of a conversation and a parallel set of affect annotations to the conversation. The affect-annotated-timeline outputs may be useful to researchers and developers, but also serve as inputs to any of a wide variety of downstream analytical processes and analysis systems that are, in turn, incorporated into many different types of special-purpose analysis and control systems.

Abstract: Aspects of the present disclosure provide systems, methods, and computer-readable storage media that support application portfolio management. Input data may be received that is associated with a set of applications. The input data may indicate, for each application of the set of applications, a name of the application and a description of the application. For an application of the set of applications, a functional score, a cost, and a technical score may be determined based on the name of the application, the description of the application, or a combination thereof. A disposition recommendation for the application may be determined based on the functional score, the cost, and the technical score. In some implementation, an indication of the disposition recommendation of the application may be output via a graphical user interface.

Abstract: The invention relates to a method of predicting a level of wear of a wear component of a vertical mill, whereby the vertical mill is used for grinding feed material. The method includes, while the mill is in operation, (i) capturing/obtaining a thermal profile of the mill from outside a grinding chamber of the mill, and (ii) determining an indication as to a level of wear of at least one wear component of the mill which is located inside the grinding chamber, based on the thermal profile of the mill. The step of capturing/obtaining a thermal profile may include performing a thermographic analysis of the mill from outside the grinding chamber of the mill. The determining step may include determining whether the wear component requires changing/replacement based on the thermographic analysis of the mill.