Abstract: A computer-implemented method for adaptively discretizing a position of a textual object in a document includes receiving, by a computer system, an image of the document and determining, by the computer system, an absolute position of the textual object in the image of the document. The method further includes normalizing, by the computer system, the absolute position to determine a relative position of the textual object. The method also includes calculating, by the computer system, a bin size such that at least one axis of the image is divided into a plurality of separate bins, wherein a distance between each bin along the at least one axis and its adjacent bin equals the bin size. The method includes discretizing, by the computer system, the relative position based on the bin size to determine a discretized position of the textual object; and providing, by the computer system, the discretized position and a textual content of the textual object as an input to a machine learning model.

Abstract: Video coding using warped motion compensation is described. Extended rotations for the warped motion compensation can be explicitly signaled. For example, motion parameters for predicting the current block and a rotation angle can be decoded. A warping matrix is obtained using the motion parameters and the rotation angle, and a prediction block is obtained by projecting the current block to a quadrilateral in a reference frame. Also described is determining a prediction model of the current block and obtaining a prediction block by projecting the current block to a quadrilateral in a reference frame. Determining the prediction model can include determining whether to predict the current block using a motion vector, a local warping model, or a global motion model, obtaining motion parameters of the prediction model, decoding a rotation angle, and obtaining a warping matrix using the motion parameters and the rotation angle.

Abstract: Methods and apparatuses for multi-document analysis and data validation during an electronic transaction workflow include a server that selects a transaction template comprising (i) data fields required to execute the transaction and (ii) business logic for validating supporting documents. The server captures digital files that contain a representation of supporting documents, identifies text in each of the captured digital files, and validates the text identified in the captured digital files using business logic. The server submits a transaction execution request including the digital files to a transaction engine and transmits a response to the transaction request to a remote computing device.

Abstract: The present disclosure relates to systems and methods for object recognition. The systems may obtain an image associated with an object. The systems may determine, using a recognition model, an initial recognition result of the object based on the image. The initial recognition result may include a plurality of initial recognition values corresponding to a plurality of predetermined categories respectively. Each of the plurality of initial recognition values may indicate an initial probability that the object corresponds to a corresponding predetermined category. In response to determining that the initial recognition result satisfies a preset condition, for each of the plurality of predetermined categories, the systems may obtain a reference probability corresponding to the predetermined category.

Abstract: Syntax elements are written to a bitstream to designate bit depth precision for palette mode coding of video blocks. During encoding, a bit depth to use for palette mode coding a current block may be based on an input video signal including the current block or based on some change in bit depth precision. A prediction residual for the current block is encoded to a bitstream along with syntax elements indicative of the bit depth used for the palette mode coding of the current block. In particular, the syntax elements include a first element indicating the palette mode coding bit depth used and a second element indicating whether to apply a bit offset to the palette mode coding bit depth. During decoding, values of the syntax elements are read from the bitstream and used to determine a bit depth for palette mode coding the encoded block.

Abstract: There is provided a method of producing Cl oxygenated products with a GaN catalyst. The Cl oxygenated products are preferably methanol and formic acid. The method is optionally selective for methanol or formic acid. The GaN catalyst and an aqueous phase are provided in a reactor having a closed environment. 0.1 to 0.55 bar of methane and from 0 to 10 bar of oxygen are introduced in the reactor. The reactor is irradiated with ultra-violet light until Cl oxygenated products such as methanol and formic acid are obtained.

Abstract: Therefore, what is needed are systems and methods for measuring the performance of a large language models (LLM). As described herein, the system generates measurement tools that are capable of accurately determining whether a predicted answer generated by an LLM is correct (in view of the corresponding question and/or reference answer). In addition, because the system does not suffer from the effects of AI hallucinations (and therefore can provide the correct determination), such determination can be performed without the need for a human to check whether the LLM is correct.

Abstract: A compressed bitstream is stored on a non-transitory computer-readable storage medium. The compressed bitstream includes an adaptive intra-prediction mode that indicates a number of filter coefficients and a subset of neighboring pixel locations relative to a pixel to be predicted. A processor processes the compressed bitstream by performing operations that include decoding the adaptive filter mode from the compressed bitstream, determining the filter coefficients for the adaptive filter mode using a training region of reconstructed pixels adjacent to a block to be decoded, and generating a prediction block for the block by recursive extrapolation. The recursive extrapolation applies the filter coefficients to pixels at the subset of neighboring pixel locations to predict each pixel of the prediction block.

Abstract: Methods and apparatus for complementary evidence identification in natural language inference. A given question is obtained and a set of N passages is obtained from a database. A probability is determined, for each passage of the set of N passages, of a corresponding passage being a supportive passage for the given question and the set of N passages is ranked based on the determined probabilities. M passages that are ranked 1 to M of the set of N passages are selected. A set of L passages is selected based on a plurality of scores, each score assigned to a set of candidate passages of the set of N passages, each score being based on the determined probabilities, the selected M passages, and a weighted regulation parameter. The set of L passages is provided to a computerized machine learning system to answer the question based on the set of L passages.

Abstract: The present invention proposes an O-band silicon-based high-speed semiconductor laser diode for optical communication and its manufacturing method, by using different buffer layers to form the growth surface of InP material with low dislocation density; N—InAlGaAs is used instead of conventional N—InAlAs electron-blocking layer in the epi-structure to reduce the barrier for electrons to enter the quantum wells from N-type and lower the threshold; a superlattice structure quantum barrier is used instead of a single layer barrier structure to improve the transport of heavy holes in the quantum wells; and the material structure is adjusted to achieve a reliable O-band high direct modulation speed semiconductor laser diode for optical communication on silicon substrate.

Abstract: A battery compartment, a battery compartment assembly and an electric scooter are provided. The battery compartment is used for an electric scooter, and includes a compartment body and a limiting flange. The electric scooter includes a frame with an opening, and the opening runs through the frame in an up-down direction. The compartment body is configured to penetrate through the opening, and the limiting flange is arranged to an outer side wall of the compartment body and configured to abut against the frame to prevent the compartment body from falling out of the opening downwards.

Abstract: A wheel hub driving system, including a housing, an electric motor, two planetary gear sets and a wheel hub bearing. The wheel hub bearing has a middle axle and an outer ring, the middle axle is fixed relative to the housing, and both a primary sun gear and a secondary sun gear are sleeved around the periphery of the wheel hub bearing. A stator of the electric motor, a primary geared ring and a secondary geared ring are all fixed to the housing. A rotor of the electric motor is anti-torsionally connected to the primary sun gear. A primary planet carrier is anti-torsionally connected to the secondary sun gear, a secondary planet carrier is anti-torsionally connected to the outer ring and the outer ring is configured to be anti-torsionally connected to a wheel rim so as to transmit power from the electric motor to the wheel rim.

Abstract: A compressed bitstream is configured for decoding by operations that include identifying an adaptive intra-prediction mode indicative of at least a training region or a configuration of neighboring pixel locations. The training region neighbors a block and consists of reconstructed pixels. Filter coefficients used to obtain respective prediction pixels of neighboring pixels within the training region when applied according to the configuration of the neighboring pixels are determined. The filter coefficients minimize a function of differences. Each difference is a respective difference between a pixel in the training region and a prediction of that pixel in the training region. A prediction block is generated for the block by recursive extrapolations that use the filter coefficients by predicting each pixel of the prediction block by applying the filter coefficients to the configuration of neighboring pixels for the pixel being predicted.

Abstract: Described herein are systems and methods for extracting borderless checkbox tables from electronic documents. A server detects checkboxes in a textual electronic document. The server extracts text blocks from the textual electronic document. The server identifies table headers corresponding to a borderless checkbox table in the textual electronic document based on the text blocks. The server determines a table boundary corresponding to the borderless checkbox table based on the table headers. The server identifies table rows and table columns corresponding to the borderless checkbox table based on the table boundary and the checkboxes. The server identifies table cells corresponding to the borderless checkbox table based on the table rows and the table columns. The server generates a data structure comprising data representing the borderless checkbox table based on at least the table cells and the checkboxes.

Abstract: Systems and methods of predicting the retention rates of students attending an educational institution are provided. The method includes receiving an initial dataset related to a cohort of students attending the educational institution, dividing the initial dataset into a training dataset and a testing dataset, training a predictive algorithm via the training dataset to generate a prediction model, and processing the testing dataset via the prediction model to output a prediction results dataset. The prediction results dataset includes a listing of the cohort of students organized from most likely to leave the educational institution to least likely to leave the educational institution. The method includes filtering a percentage of the prediction results dataset to identify a watchlist of students likely to leave the educational institution.

Abstract: A computer-implemented method for downscaling an ocean surface wind (OSW). The method comprises: obtaining an input image with a set of wind data, constructing an OSW downscaling model based on a TransUNet architecture; using the OSW downscaling model to generate an output image with a set of resulting wind data according to the input image, and before the OSW downscaling model is used to generate the output image, training the OSW downscaling model according to a plurality of sets of training samples, wherein the OSW downscaling model is trained by optimizing model parameters of the OSW downscaling model in a sense of minimizing a loss function for improving resolution.

Abstract: Remote sensing satellite data services are limited by lack of data processing and analytic capabilities as current solutions rely on manual interventions, which are inefficient, cost prohibitive for large-scale processing, and prone to human errors. A system that utilizes a large language model to understand user intent and provokes corresponding computer vision models fine-tuned with remote sensing imagery datasets, such as open vocabulary object detection and segmentation model with state-of-the-art model architecture, is provided to enable users to extract useful insights by natural language text query.

Abstract: A computer-implemented method for adaptively discretizing a position of a textual object in a document includes receiving, by a computer system, an image of the document and determining, by the computer system, an absolute position of the textual object in the image of the document. The method further includes normalizing, by the computer system, the absolute position to determine a relative position of the textual object. The method also includes calculating, by the computer system, a bin size such that at least one axis of the image is divided into a plurality of separate bins, wherein a distance between each bin along the at least one axis and its adjacent bin equals the bin size. The method includes discretizing, by the computer system, the relative position based on the bin size to determine a discretized position of the textual object; and providing, by the computer system, the discretized position and a textual content of the textual object as an input to a machine learning model.

Abstract: Systems and methods are described for enhancing performance of a search engine by using pre-existing fine-tuned machine learning models that include domain-specific knowledge. A computing device receives search results from a search engine based on a query, in which the search results include chunk identifiers and a first weight score associated with each chunk identifier. The computing device further determines a subset of search results, which includes a first predetermined number of top-ranked chunk identifiers in the search results. The computing device generates, for each chunk identifier in the subset of search results, one or more second weight scores. Then the computing device generates, via a machine learning model, ensemble scores for the chunk identifiers. Afterwards, the computing device determines a second predetermined number of top-ranked chunk identifiers based on the ensemble scores.

Abstract: For a coding block of an image, a luma prediction block is generated, a luma residual block is generated, a quantized luma block is generated after transforming the luma residual block using a luma transform type, and the quantized luma block is entropy encoded. A chroma prediction block is generated, a chroma residual block is generated, an initial chroma transform type for the chroma residual block is determined as the luma transform type, a quantized chroma block is generated using the chroma residual block transformed by a final chroma transform type, and the quantized chroma block is entropy encoded. When the initial chroma transform type is other than a default transform type, the final chroma transform type is the initial chroma transform type or the default transform type, and quantized coefficients of the quantized chroma block depend upon quantized coefficients of the quantized luma block.