Abstract: A method and apparatus for training a learning model for automatic defect detection and classification of at least a portion of a processed wafer include receiving labeled images having defect classification types and features for portions of a post-processed wafer, creating a first training set comprising the received labeled images, training the machine learning model to automatically classify wafer portions based on at least one detected defect in respective wafer portions using the first training set, receiving labeled wafer profiles having respective downstream yield data, creating a second training set comprising the labeled wafer profiles and training the machine learning model, using the second training set, to automatically determine a respective downstream yield of a wafer based on a respective wafer profile. The machine learning model can be applied to at least one unlabeled wafer image to determine at least one defect classification for at least one portion of a wafer.

Abstract: A method, apparatus and system for the automatic detection and measurement of chipping defects on diced wafers includes receiving an image of at least a portion of a diced wafer, aligning the received image of the at least the portion of the diced wafer, determining edges of the at least the portion of the diced wafer depicted in the aligned, received image, automatically determining at least one baseline from which to measure chipping defects on the at least the portion of the diced wafer from the determined edges, and measuring chipping defects on the at least the portion of the diced wafer using at least one determined, respective baseline. In some embodiments, the method, apparatus and system can further include applying a machine learning model to measured chipping defects to determine if a critical failure exists on the diced wafer.

Abstract: A method and apparatus for training a learning model for the automatic detection and classification of defects on wafers includes receiving labeled images of wafer defects having multiple defect classifications, creating a first training set including the received labeled images of wafer defects, training the machine learning model to automatically detect and classify wafer defects in a first stage using the first training set, blending at least one set of at least two labeled images having different classifications to generate additional labeled image data, creating a second training set including the blended, additional labeled image data, and training the machine learning model to automatically detect and classify wafer defects in a second stage using the second training set. The trained machine learning model can then be applied to at least one unlabeled wafer image to determine at least one defect classification for the at least one unlabeled wafer image.

Abstract: A semiconductor device includes: a substrate comprising a magnetic tunneling junction (MTJ) region and a logic region; a first MTJ on the MTJ region; a first metal interconnection on the logic region; and a cap layer extending from a sidewall of the first MTJ to a sidewall of the first metal interconnection. Preferably, the cap layer on the MTJ region and the cap layer on the logic region comprise different thicknesses.

Abstract: The present invention relates to an extension structure of flexible substrates with conductive wires thereon. In a first embodiment, three flexible substrates are prepared, each having multiple conductive wires configured on their front surfaces. The third flexible substrate is flipped over, with its conductive wires facing downwards, and bonded across a boundary formed by the first and second flexible substrates. As a result, the corresponding conductive wires between the first and second flexible substrates are electrically coupled with each other through being physically pressed by corresponding conductive wires in the third flexible substrate.

Abstract: A chemical mechanical polishing (CMP) apparatus is provided, including a polishing pad and a polishing head. The polishing pad has a polishing surface. The polishing head is configured to hold a wafer in contact with the polishing surface during the polishing process. The polishing head includes a retaining ring, at least one fluid channel, and a vacuum pump. The retaining ring is arranged along the periphery of the polishing head and configured to retain the wafer. The at least one fluid channel is provided inside the polishing head, wherein the retaining ring includes a bottom surface facing the polishing surface and a plurality of holes in fluid communication with the bottom surface and the at least one fluid channel. The vacuum pump is fluidly coupled to the at least one fluid channel.

Abstract: A sequence-to-sequence summarizer receives source content to be summarized and determines whether the source content has a size that meets the size threshold. If so, the source content is divided into sections and the sequence-to-sequence summarizer generates a summary for each section. The summaries for each section are merged into a document summary and surfaced for user interaction.

Abstract: A charging circuitry includes a power electronic converter, a current sensor, a voltage boost/buck controller and a charging mode controller. The power electronic converter is configured to charge or discharge a supercapacitor according to a control command. The current sensor is coupled to the supercapacitor for detecting a first sensed voltage and a second sensed voltage. The voltage boost/buck controller is configured to generate the control command and a current command according to the first and second sensed voltages and an overall feedback. The charging mode controller is configured to generate a current feedback and a voltage feedback to the voltage boost/buck controller according to a driving voltage, the current command and a third sensed voltage of the supercapacitor. The third sensed voltage, the current feedback and the voltage feedback are superposed as the overall feedback and then inputted to the same input terminal of the voltage boost/buck converter.

Abstract: Disclosed are a method and system for intelligent source tracing of organic pollution of a water body, which belongs to the technical field of environmental analytical chemistry. The method comprises: acquiring organic matter analysis and detection data from water samples of a polluted water body through high performance liquid chromatography-tandem mass spectrometry; performing high-throughput screening on the organic matter in the water samples according to said data to determine pollutants; identifying pollution sources by means of network analysis according to the determined pollutants; and according to the identified pollution sources and the organic pollutants, determining key pollutants and quantifying the pollution contributions thereof by using a machine learning classification model.

Abstract: An electronic transformer includes a first forward rectifier, a second forward rectifier, a third forward rectifier and a backward rectifier. The first forward rectifier is coupled between a first-phase power and a first output terminal. The second forward rectifier is coupled between a second-phase power and the first output terminal. The third forward rectifier is coupled between a third-phase power and the first output terminal. The backward rectifier is coupled between a neutral line and a second output terminal. The first forward rectifier, the second forward rectifier, and the third forward rectifier are configured to half-wave rectify the first-phase power, the second-phase power, and the third-phase power to generate rectified first-phase to third-phase power sources, and superimpose the rectified first-phase to third-phase power sources on the first output end to serve as an output voltage of the electronic transformer.

Abstract: A portable charging case, including an unlocking assembly and a push rod, user can push the unlocking assembly in the direction towards the storage space after the portable lighting device is fully charged. The push rod then pushes the magnetic charging device away from the portable lighting device so that the magnetic charging device disconnects from the portable lighting device. Users can take the portable lighting device out from the charging case more quickly and easily, thereby enhancing user experience for using the charging case.

Abstract: Behavior report generation monitors the behavior of unknown sample files executing in a sandbox. Behaviors are encoded and feature vectors created based upon a q-gram for each sample. Prototypes extraction includes extracting prototypes from the training set of feature vectors using a clustering algorithm. Once prototypes are identified in this training process, the prototypes with unknown labels are reviewed by domain experts who add a label to each prototype. A K-Nearest Neighbor Graph is used to merge prototypes into fewer prototypes without using a fixed distance threshold and then assigning a malware family name to each remaining prototype. An input unknown sample can be classified using the remaining prototypes and using a fixed distance. For the case that no such prototype is close enough, the behavior report of a sample is rejected and tagged as an unknown sample or that of an emerging malware family.

Abstract: Disclosed are a method and system for intelligent source tracing of organic pollution of a water body, which belongs to the technical field of environmental analytical chemistry. The method comprises: acquiring organic matter analysis and detection data from water samples of a polluted water body through high performance liquid chromatography-tandem mass spectrometry; performing high-throughput screening on the organic matter in the water samples according to said data to determine pollutants; identifying pollution sources by means of network analysis according to the determined pollutants; and according to the identified pollution sources and the organic pollutants, determining key pollutants and quantifying the pollution contributions thereof by using a machine learning classification model.

Abstract: The present disclosure relates to a photosensitive structure and a mobile lighting device. The photosensitive structure includes a head housing, and a metal substrate, a lens, a shading bowl, and a light sensor that are located in the head housing. The metal substrate, the lens, and the light sensor are all fixed in the head housing, and the metal substrate is located between the lens and the light sensor, the metal substrate is electrically connected to the light sensor. The shading bowl is fixed to the metal substrate and located between the metal substrate and the lens. Along a light incoming direction, the photosensitive structure is provided with a light inlet located outside a projection of the shading bowl on the metal substrate, the light sensor is configured to sense an external brightness through the light inlet.

Abstract: A discourse-level text translation method and device, the method comprising: acquiring a text to be translated, the text to be translated being a unit text in a discourse-level text to be translated (S101); acquiring an associated text of the text to be translated, the associated text including at least one of a preceding source text, a following source text, and a preceding target text (S102); and translating, according to the associated text, the text to be translated (S103).

Abstract: A speech recognition method and related products are provided. The method includes acquiring a to-be-recognized speech and a configured hot word library; determining, based on the to-be-recognized speech and the hot word library, an audio-related feature used at a current decoding time instant; determining, based on the audio-related feature, a hot word-related feature used at the current decoding time instant from the hot word library; and determining, based on the audio-related feature and the hot word-related feature, a recognition result of the to-be-recognized speech at the current decoding time instant.

Abstract: A method and a system for end-to-end modeling are provided. The method includes: determining a topological structure of a target-based end-to-end model, where the topological structure includes an input layer, an encoding layer, an code enhancement layer, a filtering layer, a decoding layer and an output layer; the code enhancement layer adds information of a target unit to a feature sequence outputted by the encoding layer, the filtering layer filters a feature sequence added with the information of the target unit; collecting multiple pieces of training data; and training parameters of the target-based end-to-end model by using the multiple pieces of the training data.

Abstract: A system and method for scheduling at least one event includes a component module arranged to receive component data regarding a plurality of possible operation decisions, a modelling module arranged to utilise the plurality of operation decisions to construct a Petri net model to simulate a real world system, a scheduling module arranged to utilise the Petri net and variable data regarding the plurality of operation decisions to determine a schedule of events that is feasible subject to a set of predetermined constraints, and an operation module arranged to provide the schedule of events to at least one external device.

Abstract: A portable charging case, including an unlocking assembly and a push rod, user can push the unlocking assembly in the direction towards the storage space after the portable lighting device is fully charged. The push rod then pushes the magnetic charging device away from the portable lighting device so that the magnetic charging device disconnects from the portable lighting device. Users can take the portable lighting device out from the charging case more quickly and easily, thereby enhancing user experience for using the charging case.

Abstract: Disclosed is a method for high-throughput screening of non-target biomarkers based on metabolic disturbance caused by pollutants, belonging to the field of environmental exposure and health.