Patents by Inventor Stefan Klaus
Stefan Klaus has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Publication number: 20240045890Abstract: Methods, systems, and computer-readable storage media for a machine learning (ML) system for matching a query entity to one or more target entities, the ML system that reducing a number of query-target entity pairs from consideration as potential matches during inference.Type: ApplicationFiled: August 4, 2022Publication date: February 8, 2024Inventors: Hoang-Vu Nguyen, Li Rong Wang, Matthias Frank, Rajesh Vellore Arumugam, Stefan Klaus Baur, Sundeep Gullapudi
-
Publication number: 20230325708Abstract: Computer-readable media, methods, and systems are disclosed for feature attribution in a machine learning model. Samples may be generated for a machine learning model based on a normalized probability distribution. The samples may be used to determine a weight for features and feature pairs for the machine learning model. The weights of the features and feature pairs may be used to determine which features are significant for predictions within the machine learning model.Type: ApplicationFiled: April 12, 2022Publication date: October 12, 2023Inventors: Stefan Klaus Baur, Matthias Frank, Hoang-Vu Nguyen, Kannan Presanna Kumar
-
Publication number: 20230222147Abstract: Methods, systems, and computer-readable storage media for receiving a set of inference results generated by a ML model, the inference results including a set of query entities and a set of target entities, each query entity having one or more target entities matched thereto by the ML model, processing the set of inference results to generate a set of matched sub-sets of target entities by executing a search over target entities in the set of target entities based on constraints, for each problem in a set of problems, providing the problem as a tuple including an index value representative of a target entity in the set of target entities and a value associated with the query entity, the value including a constraint relative to the query entity, and executing at least one task in response to one or more matched sub-sets in the set of matched sub-sets.Type: ApplicationFiled: January 10, 2022Publication date: July 13, 2023Inventors: Hoang-Vu Nguyen, Rajesh Vellore Arumugam, Matthias Frank, Stefan Klaus Baur
-
Patent number: 11687575Abstract: Methods, systems, and computer-readable storage media for receiving a set of inference results generated by a ML model, the inference results including a set of query entities and a set of target entities, each query entity having one or more target entities matched thereto by the ML model, processing the set of inference results to generate a set of matched sub-sets of target entities by executing a search over target entities in the set of target entities based on constraints, for each problem in a set of problems, providing the problem as a tuple including an index value representative of a target entity in the set of target entities and a value associated with the query entity, the value including a constraint relative to the query entity, and executing at least one task in response to one or more matched sub-sets in the set of matched sub-sets.Type: GrantFiled: January 10, 2022Date of Patent: June 27, 2023Assignee: SAP SEInventors: Hoang-Vu Nguyen, Rajesh Vellore Arumugam, Matthias Frank, Stefan Klaus Baur
-
Publication number: 20230128485Abstract: Methods, systems, and computer-readable storage media for receiving IRF data sets, the IRF data sets including a set of records including inference results determined by the ML model during production use of the ML model and at least one correction to an inference result, executing incremental training of the ML model to provide an updated ML model by selectively filtering one or more records of the set of records to adjust a negative sample to positive sample proportion of a sub-set of records based on a negative sample to positive sample proportion of initial training of the ML model, for each record in the sub-set of records, determining a weight, and during incremental training, applying the weight of a respective record being in a loss function in determining an accuracy of the ML model based on the respective record, and deploying the updated ML model for production use.Type: ApplicationFiled: October 27, 2021Publication date: April 27, 2023Inventors: Rajesh Vellore Arumugam, Andrew Ivan Soegeng, Stefan Klaus Baur
-
Patent number: 11615120Abstract: Pairwise entity matching systems and methods are disclosed herein. A deep learning model may be used to match entities from separate data tables. Entities may be preprocessed to fuse textual and numeric data early in the neural network architecture. Numeric data may be represented as a vector of a geometrically progressing function. By fusing textual and numeric data, including dates, early in the neural network architecture the neural network may better learn the relationships between the numeric and textual data. Once preprocessed, the paired entities may be scored and matched using a neural network.Type: GrantFiled: July 14, 2021Date of Patent: March 28, 2023Assignee: SAP SEInventors: Stefan Klaus Baur, Matthias Frank, Hoang-Vu Nguyen
-
Publication number: 20220391414Abstract: Pairwise entity matching systems and methods are disclosed herein. A deep learning model may be used to match entities from separate data tables. Entities may be preprocessed to fuse textual and numeric data early in the neural network architecture. Numeric data may be represented as a vector of a geometrically progressing function. By fusing textual and numeric data, including dates, early in the neural network architecture the neural network may better learn the relationships between the numeric and textual data. Once preprocessed, the paired entities may be scored and matched using a neural network.Type: ApplicationFiled: July 14, 2021Publication date: December 8, 2022Inventors: Stefan Klaus Baur, Matthias Frank, Hoang-Vu Nguyen
-
Patent number: 11470964Abstract: A piece of furniture or domestic appliance includes a body, a drawer element slidably arranged in the body and having two side walls. Respective pull-out guides are arranged between the side panels and side walls of the body. Each of the pull-out guides has at least one body rail fastenable to a side wall of the body and one running rail fastenable to the side panel. The piece of furniture or domestic appliance also includes at least one functional unit, which is coupled to one of the pull-out guides. A functional unit is detachably arranged in at least one of the side walls at distance from the pull-out guide, which functional unit, in a predefined position of the drawer element relative to the body, is operatively connected to a counter-element stationarily mounted on the body or on the body rail.Type: GrantFiled: February 8, 2019Date of Patent: October 18, 2022Assignee: PAUL HETTICH GMBH & CO. KGInventors: Timo Szlapka, Sebastian Bastkowski, Thorsten Hötger, Gerrit Staufenberg, Ulrich Ratsch, Stefan Klaus
-
Patent number: 11464331Abstract: A pull-out guide has at least two rails that are movable relative to each other. A first attaching means for attaching a first pushing element is provided on a movable rail. A second attaching means at a distance from the first attaching means is provided on the movable rail for attaching a pushing element of a second drawer system. The pushing element of either the first or second drawer system can be attached to the movable rail.Type: GrantFiled: March 22, 2019Date of Patent: October 11, 2022Assignee: Paul Hettich GmbH & Co. KGInventor: Stefan Klaus
-
Patent number: 11312424Abstract: A vehicle load-bearing structure includes a cell made of composite materials and composed of a plurality of components joined to each other by at least one joining system to form a single structure; the cell is made up of a number n of components between 3 and 8, that is 3<n<8.Type: GrantFiled: December 13, 2017Date of Patent: April 26, 2022Assignee: Automobili Lamborghini S.P.A.Inventors: Stefan Klaus Nothdurfter, Andrea Aguggiaro, Karsten Schuffenhauer
-
Publication number: 20220092405Abstract: In an example embodiment, a deep neural network may be utilized to determine matches between candidate pairs of entities, as well as confidence scores that reflect how certain the deep neural network is about the corresponding match. The deep neural network is also able to find these matches without requiring domain knowledge that would be required if features for a machine-learned model were handcrafted, which is a drawback of prior art machine-learned models used to match entities in multiple tables. Thus, the deep neural network improves on the functioning of prior art machine learned models designed to perform the same tasks. Specifically, the deep neural network learns the relationships of tabular fields and the patterns that define a match from historical data alone, making this approach generic and applicable independent of the context.Type: ApplicationFiled: September 18, 2020Publication date: March 24, 2022Inventors: Matthias Frank, Hoang-Vu Nguyen, Stefan Klaus Baur, Alexey Streltsov, Jasmin Mankad, Cordula Guder, Konrad Schenk, Philipp Lukas Jamscikov, Rohit Kumar Gupta
-
Patent number: 11275934Abstract: Disclosed herein are system, method, and computer program product embodiments for generating document labels using positional embeddings. In an embodiment, a label system may identify tokens, such as words, of a document image. The label system may apply a position vector neural network to the document image to analyze the pixels and determine positional embedding vectors corresponding to the words. The label system may then combine the positional embedding vectors to corresponding word vectors for use as an input to a neural network trained to generate document labels. This combination may embed the positional information with the corresponding word information in a serialized manner for processing by the document label neural network. Using this formatting, the label system may generate document labels in a light-weight and fast manner while still preserving spatial relationships between words.Type: GrantFiled: November 20, 2019Date of Patent: March 15, 2022Assignee: SAP SEInventors: Christian Reisswig, Stefan Klaus Baur
-
Publication number: 20210204136Abstract: A system includes a plurality of antennas, an access control mechanism, and a computing resource. The computing resource is configured to initiate each of multiple antennas to transmit a wireless signal and receive values indicative of signal strength of the wireless signals from the multiple antennas. The computing resource also is configured to calculate a position of a wireless electrical device based on the received values and calculate an error value of the calculated position of the wireless electrical device. Further, the computing resource is configured to determine that the error value is greater than an error threshold and to disable the access control mechanism.Type: ApplicationFiled: January 15, 2020Publication date: July 1, 2021Inventors: Werner Lummer, Stefan Klaus Beierke
-
Publication number: 20210150201Abstract: Disclosed herein are system, method, and computer program product embodiments for generating document labels using positional embeddings. In an embodiment, a label system may identify tokens, such as words, of a document image. The label system may apply a position vector neural network to the document image to analyze the pixels and determine positional embedding vectors corresponding to the words. The label system may then combine the positional embedding vectors to corresponding word vectors for use as an input to a neural network trained to generate document labels. This combination may embed the positional information with the corresponding word information in a serialized manner for processing by the document label neural network. Using this formatting, the label system may generate document labels in a light-weight and fast manner while still preserving spatial relationships between words.Type: ApplicationFiled: November 20, 2019Publication date: May 20, 2021Inventors: Christian Reisswig, Stefan Klaus Baur
-
Publication number: 20210015257Abstract: A pull-out guide has at least two rails that are movable relative to each other. A first attaching means for attaching a first pushing element is provided on a movable rail. A second attaching means at a distance from the first attaching means is provided on the movable rail for attaching a pushing element of a second drawer system. The pushing element of either the first or second drawer system can be attached to the movable rail.Type: ApplicationFiled: March 22, 2019Publication date: January 21, 2021Applicant: Paul Hettich GmbH & Co. KGInventor: Stefan KLAUS
-
Publication number: 20200245819Abstract: A device for fastening a handle to cookware or bakeware having a support piece protruding outward from the cookware or bakeware wall. The handle is detachably snapped onto the support piece, and has a frontal recess, and a spring element retained in the recess acting on a latching element of the. When the handle is snapped onto the support piece, an actuation element activates the spring element. The support piece has a fork element clamped in the recess of the handle at the surface parallel to its base and is supported on a support surface in an operating position.Type: ApplicationFiled: January 30, 2020Publication date: August 6, 2020Inventors: Michael Besche, Stefan Klaus
-
Publication number: 20200069114Abstract: A fitting part for attaching a handle element to the vessel wall of a cooking or frying vessel includes a base body having a recess for plugging onto a support part on the vessel wall, a securing element for securing the base body to the support part provided in the recess, and the securing element engaging at least one cast surface of the support part and/or at least one cast surface of the base body in a fixing manner, the securing element being at least one elastically deformable pressure spring having a wavelike profile, the opposite sides of the of the wavelike profile coming into contact with the support part on one side and with the inner side of the recess in the base body on the other side, and the pressure spring fixing the base body to the support part by friction-locking contact with at least one cast surface.Type: ApplicationFiled: August 30, 2019Publication date: March 5, 2020Inventors: Michael Besche, Stefan Klaus
-
Publication number: 20190283810Abstract: A vehicle load-bearing structure includes a cell made of composite materials and composed of a plurality of components joined to each other by at least one joining system to form a single structure; the cell is made up of a number n of components between 3 and 8, that is 3<n<8.Type: ApplicationFiled: December 13, 2017Publication date: September 19, 2019Inventors: Stefan Klaus NOTHDURFTER, Andrea AGUGGIARO, Karsten SCHUFFENHAUER
-
Patent number: 10085558Abstract: A pull-out guide for a drawer, comprising a guide rail can be fixed to a body and a running rail which is movably mounted. At least one central rail can optionally be provided between the guide rail and the running rail, and a securing device with a holder is provided on the running rail, wherein a drawer can be fixed to the holder. The holder is mounted in a movable manner in the longitudinal direction of the running rail and is pretensioned towards the front in the pull-out direction by means of an energy accumulator. In this manner, a drawer can be easily mounted on the securing device, and manufacturing tolerances can be compensated for.Type: GrantFiled: March 25, 2015Date of Patent: October 2, 2018Assignee: PAUL HETTICH GMBH & CO. KGInventor: Stefan Klaus
-
Publication number: 20170105525Abstract: A pull-out guide for a drawer, comprising a guide rail can be fixed to a body and a running rail which is movably mounted. At least one central rail can optionally be provided between the guide rail and the running rail, and a securing device with a holder is provided on the running rail, wherein a drawer can be fixed to the holder. The holder is mounted in a movable manner in the longitudinal direction of the running rail and is pretensioned towards the front in the pull-out direction by means of an energy accumulator. In this manner, a drawer can be easily mounted on the securing device, and manufacturing tolerances can be compensated for.Type: ApplicationFiled: March 25, 2015Publication date: April 20, 2017Inventor: Stefan KLAUS