Abstract: Methods, systems, and articles of manufacture, including computer program products, are provided for synthesizing images for machine learning. The method may include selecting one or more image preprocessing transformations to apply on the foreground object image; applying the selected one or more image preprocessing transformations to the foreground object image; selecting a background image from a set of background images depicting a variety of different backgrounds which may be associated with the foreground object image; merging the selected background image with the foreground object image to form a synthesized image; selecting one or more image transformations to apply on the synthesized image; applying the selected one or more image transformations to the synthesized image; and storing the synthesized image in a collection of synthesized images to train a machine learning model.

Abstract: A server chassis includes a main frame, a fan module, and a flow-guiding device. The fan module and the flow-guiding device are in the main frame. The flow-guiding device includes a plurality of plate bodies. The plate bodies are parallel to each other and are spaced apart from each other by a spacing. Each of the plate bodies has a plurality of openings. The plurality of openings of one of the plate bodies are arranged alternately with those of another plate body adjacent to the one of the plate bodies.

Abstract: Disclosed herein are various embodiments for targeted document information extraction. An embodiment operates by receiving a document associated with a particular customer of a plurality of customers. It is determined whether to use a global processor or template processor to analyze the document based on whether one or more customer templates are associated with the particular customer. Which of the one or more templates associated with the particular customer correspond to the document is identified. The document is compared to the identified template associated with the customer. Information is extracted from the document based on the identified template and the identified plurality of variations. The extracted information for the document is output.

Abstract: A system for neuromodulation includes a split-ring resonator (SRR) comprising a resonance circuit, the SRR being implantable in a cranial target site and a source of microwave signals, wherein the microwave signals are deliverable wirelessly to couple with the SRR to produce a localized electrical field, wherein the localized electrical field inhibits one or more neurons at the cranial target site with submillimeter spatial precision.

Abstract: In some embodiments, there is provided a system. The system may include at least one data processor and at least one memory storing instructions which, when executed by the at least one data processor, cause the apparatus to at least: determine, for a received document including at least one item, that the received document likely includes at least one missing item, the determination based on at least a machine learning model and the at least one item; and provide an indication of the at least one missing item. Related systems and articles of manufacture are also provided.

Abstract: A first converter station is part of a high voltage direct current transmission system that includes a DC transmission link connected to the first converter station a second converter station. A DC current and a DC voltage of the DC transmission link are sensed by the first converter station. It is determined that the sensed DC current is equal to or larger than a threshold current value, that the sensed DC current is equal to or larger than the threshold current value, and that at least a partial recovery of the sensed DC voltage has occurred. On the basis that it is determined that the at least a partial recovery of the sensed DC voltage has occurred, it is determined that a phase-to-ground fault has occurred. In response to determining that a phase-to-ground fault has occurred, a power delivered by the first converter station can be reduced.

Abstract: A tapered fiber optoacoustic emitter includes a nanosecond laser configured to emit laser pulses and an optic fiber. The optic fiber includes a tip configured to guide the laser pulses. The tip has a coating including a diffusion layer and a thermal expansion layer, wherein the diffusion layer includes epoxy and zinc oxide nanoparticles configured to diffuse the light while restricting localized heating. The thermal expansion layer includes carbon nanotubes (CNTs) and Polydimethylsiloxane (PDMS) configured to convert the laser pulses to generate ultrasound. The frequency of the ultrasound is tuned with a thickness of the diffusion layer and a CNT concentration of the expansion layer.

Abstract: Provided is a non-belt automatic mechanized sampling system for truck, including: a working platform including a transport channel and a slide rail extending along a first direction, the transport channel is arranged under the slide rail and makes the truck pass; a first frame configured to slide with the slide rail; a sampling mechanism (configured to collect materials from the truck along a second direction perpendicular to the first direction, the first and second directions are arranged horizontally; an integrated sample preparation component including a discharging mechanism and a feeder, a crusher, a constant mass dividing machine and a sample retention barrel. The feeder transports the material to the crusher. The crusher crushes the material and transports the crushed material to the constant mass dividing machine. The constant mass dividing machine divides the crushed material into samples to the sample retention barrel, and the discarded material to the discharging mechanism.

Abstract: A server chassis includes a main frame, a fan module, and a flow-guiding device. The fan module and the flow-guiding device are in the main frame. The flow-guiding device includes a plurality of plate bodies. The plate bodies are parallel to each other and are spaced apart from each other by a spacing. Each of the plate bodies has a plurality of openings. The plurality of openings of one of the plate bodies are arranged alternately with those of another plate body adjacent to the one of the plate bodies.

Abstract: Disclosed herein are various embodiments for targeted document information extraction. An embodiment operates by receiving a document associated with a particular customer of a plurality of customers. It is determined whether to use a global processor or template processor to analyze the document based on whether one or more customer templates are associated with the particular customer. Which of the one or more templates associated with the particular customer correspond to the document is identified. The document is compared to the identified template associated with the customer. Information is extracted from the document based on the identified template and the identified plurality of variations. The extracted information for the document is output.

Abstract: A neutral arrangement is provided for a converter station of a direct current power transmission system that includes a first converter. The neutral arrangement includes surge arrestors and a group of neutral buses. Each surge arrestor is connected between a neutral bus and ground. The neutral arrangement includes a high voltage insulation zoom area having a first group of surge arrestors and a low voltage insulation zoom area having a second group of surge arrestors. The surge arrestors in the first group have a first arrestor reference voltage and the surge arrestors in the second group have a second arrestor reference voltage that is lower than the first arrestor reference voltage.

Abstract: Methods of energizing converters, and converter stations thereof, based on a mix of half-bridge (HB) converter cells and full-bridge (FB) converter cells are provided. According to one embodiment, a method for energizing a converter may include establishing a direct electrical connection between DC terminals of a converter, electrically connecting the converter to an AC power source via AC connectors and obtaining, from the AC power source and for a certain period of time, an AC voltage between two electrical phase lines for charging capacitors of the FB cells connected to the two electrical phase lines. Accordingly, the FB cells will be charged to a higher value and they may be controlled for charging capacitors of the HB submodules to at least a minimum voltage required to start switching devices of the HB submodules by drawing power from the AC power source.

Abstract: Provided is a non-belt automatic mechanized sampling system for a train, including: a working platform provided with a slide rail; a sampling mechanism slidably arranged on the slide rail and configured to collect materials from the carriage of the train; an integrated sample preparation component including a discharging mechanism and a feeder, a crusher, a constant mass dividing machine and a sample retention barrel arranged from top to bottom. The feeder transports the material to the crusher. The crusher crushes the material and transports the crushed material to the constant mass dividing machine. The constant mass dividing machine divides the crushed material into samples to the sample retention barrel, and the discarded material to the discharging mechanism. The discharging mechanism transports the discarded materials to the carriage.

Abstract: A method for forming a semiconductor structure in provided. The method includes providing a substrate, forming a gate electrode layer on the substrate, and performing a defluorination treatment on the gate electrode layer. The method also includes, after performing the defluorination treatment, forming a barrier layer on a portion of a surface of the gate electrode layer. The barrier layer is made of a material including titanium element.

Abstract: A neutral-point voltage balance control method and system for a three-level converter in a full power factor range. The method includes: using a large, medium, and zero vector modulation method to synthesize a reference voltage vector, and duty cycles of a large vector, a medium vector, and a zero vector; obtaining a voltage difference between two dc-link capacitors of signal acquisition, and the voltage difference as a neutral-point potential of the three-level converter; according to a value relationship between the neutral-point potential of the three-level converter and a specified threshold, selecting a small vector and calculating a duty cycle of the small vector; and updating a duty cycle of each basic vector, and obtaining a switching sequence for controlling a power switch of a three-phase bridge arm. An amplitude of a common-mode voltage of an NPC three-level converter is equal to one sixth of a dc-link voltage.

Abstract: The invention discloses a distributed IoT terminal system and method based on a fiber optic bus RoF. The system comprises a core application part, a pre-channel part, a transmission channel part, an acquisition field part, an acquisition terminal part and a data source. The invention adopts the principle of light-borne ROF to adapt to pan-socialization, decentralization, verticalization, refinement, full traceability, full lifecycle management and governance, instant, real-time, online, interactive and distributed IoT applications. The integrated and integrated application of the invention improves social efficiency and saves social cost. It may be integrated and integrated the application into any application systems, and realized dynamic instant, real-time, online, centimeter level positioning application. It can infer the specific location of the positioning device.

Abstract: A use of a substance for inhibiting SOAT1 gene expression and/or protein activity. The use is selected from at least one of: (a) Preparation of kits for liver cancer diagnosis; (b) Preparation of kits for liver cancer prognosis; (c) Preparation of companion diagnostic kits for treatment of liver cancer; (d) For the preparation of drugs for the prevention and/or treatment of cancer; (e) For the preparation of drugs for the prevention and/treatment of cancer spread and metastasis; (f) For the preparation of drugs that promote the apoptosis of cancer cells; (g) For the preparation of drugs for inhibiting cancer cell formation; (h) For the preparation of drugs that inhibit the proliferation and growth of cancer cells in vitro. Experiments have shown that SOAT1 is highly expressed in liver cancer tissues and serum, and its high abundance indicates poor prognosis of liver cancer patients.

Abstract: Methods, systems, and articles of manufacture, including computer program products, are provided for synthesizing images for machine learning. The method may include selecting one or more image preprocessing transformations to apply on the foreground object image; applying the selected one or more image preprocessing transformations to the foreground object image; selecting a background image from a set of background images depicting a variety of different backgrounds which may be associated with the foreground object image; merging the selected background image with the foreground object image to form a synthesized image; selecting one or more image transformations to apply on the synthesized image; applying the selected one or more image transformations to the synthesized image; and storing the synthesized image in a collection of synthesized images to train a machine learning model.