Abstract: A computer-implemented method and system relate to an image encoder that receives a digital image as input. The image encoder generates a weight map using a preceding feature map. The preceding feature map is generated using pixels of the digital image. The weight map is generated based on lie data associated with the digital image. A homographic transformation is interpolated between two planar projections of the digital image using at least the weight map and a homography matrix. The homography matrix provides a mapping between the two planar projections of the digital image. Homographic transformed kernels are generated by applying the homographic transformation to convolution kernels. The homographic transformed kernels are applied to the preceding feature map to conduct convolution on different plane regions appearing in the digital image and generate a new feature map, which is used for a computer vision task involving three-dimensional (3D) perception.

Abstract: A method includes identifying an output image of a substrate defect based on a user sketch of the substrate defect and a text description associated with the substrate defect. The method further includes updating the output image based on user input to generate an updated output image. The method further includes causing, based on the updated output image, performance of a corrective action associated with substrate processing via a substrate processing system.

Abstract: In an example embodiment, a rule repository is stored within a rules framework separate and distinct from the Case Management Model and Notation (CMMN) framework. The rule repository contains rules for evaluating and producing results from conditions within CMMN sentries. A rules runtime can then be invoked by a sentry evaluator within the CMMN framework when a sentry is encountered during a CMMN flow. The rules runtime can then retrieve and execute the corresponding BPMN rule to perform dynamic evaluation of the conditions. This design allows for the conditions to be updated without the need to update the CMMN model.

Abstract: A method includes obtaining, by a processing device, defect data for a substrate processed in a substrate processing system. The method further includes obtaining, by the processing device, context data associated with the substrate. The method further includes determining a troubleshooting guide associated with the defect data. The troubleshooting guide includes a sequence of troubleshooting operations, each associated with one or more probably root causes for the defect data. The method further includes determining a subset of context data based on the troubleshooting guide. The method further includes processing the defect data and the subset of context data using one or more trained machine learning models that output a predicted corrective action associated with a troubleshooting operation in the sequence of troubleshooting operations. The method further includes initiating the corrective action.

Abstract: A computer-implemented method and system relate to improved object detection via a machine learning system, which includes at least an image encoder, a semantic segmentation head, and an object detection head. This machine learning system exhibits improved effectiveness in detecting relatively large objects. The image encoder generates image embedding data using at least one digital image. A bird's eye view (BEV) feature map is generated using the image embedding data. The semantic segmentation head generates semantic segmentation data using the BEV feature map. The object detection head generates three-dimensional (3D) box data for a detected object of the digital image based on the BEV feature map and the semantic segmentation data. The object detection head and the semantic segmentation head are jointly trained using a combined loss, which includes a first loss based on the BEV semantic segmentation data and a second loss based on the 3D box data.

Abstract: A method includes obtaining, by a processing device, a first image of a first substrate including a first set of substrate structures. The method further includes obtaining a first set of measurements in association with the first set of substrate structures. The method further includes obtaining a second image of a second substrate including a second set of substrate structures corresponding to the first set of substrate structures. The method further includes determining a set of unit structures of the second image. Each of the unit structures is associated with one of the second set of substrate structures. Determining the set of unit structures is performed based on the first set of measurements. The method further includes determining a second set of measurements. The second set of measurements correspond to the first set of measurements. The second set of measurements are in association with the set of unit structures.

Abstract: A method includes obtaining defect data and context data in association with a substrate, and providing the defect data and the context data to a first trained machine learning model as input. The method further includes obtaining output from the first trained machine learning model based on the defect data and the context data. The output is indicative of a predicted root cause in association with the defect data. The method further includes performing a corrective action in view of the output.

Abstract: Embodiments described herein relate to a method that includes implementing a feature extraction process and a feature fusion process from a data set that includes one or more chamber setting data points, where the data set is augmented by a physics attributes model that uses the one or more chamber setting data points to generate chamber attribute data of one or more processing characteristics within a chamber based on physics modeling. In an embodiment, the method further includes implementing a data segmentation process on the data set with a context specific data segmentation module to form a modified data set. In an embodiment, the method may further include training a machine learning model on the modified data set, wherein training the machine learning model includes minimizing a loss function that includes a regularized objective function that includes a term based on physics informed variables.

Abstract: Described herein are techniques for generating recommendations for plan items that are configured for manual activation. Sentries are conditions attached to a plan item such as events, tasks, or stages within a case plan. A plurality of ML models may be trained with historical case data from an unstructured case management model. The plurality of ML models may generate recommendations when presented with current case data. The recommendations may be utilized to automatically activate the manually activated plan items when the confidence score is above a predefined threshold. Advantages of these techniques include reducing human error and fatigue as fewer decisions need to be made by humans.

Abstract: A method includes identifying substrate images that have been sorted into classes. The method further includes training a machine learning model using data input including the substrate images and target output including the classes. The method further includes refining the trained machine learning model using a triplet loss function based on one or more substrate images misclassified by the trained machine learning model to provide a refined trained machine learning model associated with performance of an action associated with substrate processing.

Abstract: An end-to-end trainable grouped mathematically differentiable non-maximal suppression (NMS) technique is presented for monocular 3D object detection. First, formulate NMS as a matrix operation and then group and mask the boxes in an unsupervised manner to obtain a simple closed-form expression of the NMS. This technique addresses the mismatch between training and inference pipelines and, therefore, forces the network to select the best 3D box in a differentiable manner. As a result, the proposed technique achieves state-of-the-art monocular 3D object detection results on the KITTI benchmark dataset performing comparably to monocular video-based methods.

Abstract: A computer-implemented user-friendly system and method of designing and managing smart contracts on a distributed ledger (blockchain). The system creates a number of computer programs that correspond to a business user's model of the contract terms. In this manner, the business user can generate the smart contract without needing to understand programming or involve third parties like developers.

Abstract: In an example embodiment, a solution for converting unstructured process models to structured process models is described. Machine learning is used to automatically create a first set of suggested optimizations for an unstructured process model based upon context data (and possibly execution logs as well) for the unstructured process model. A heuristic rules engine can then create a second set of suggested optimizations for the unstructured process model. The suggested optimizations may be presented to a user via a user interface. Some or all of the optimizations may then be performed on the unstructured process model, and then the optimized unstructured process model can be converted to a structured process model.

Abstract: The present disclosure relates to synthetic biology and, in particular, the bioproduction of bakuchiol, and engineered enzymes for producing the same.

Abstract: A metaverse environment reader performs semantic segmentation and object detection steps to identify a plurality of objects in a metaverse scene. Next, the reader determines an order of importance of the plurality of objects in the scene based at least on a location and a size of each object. Then, the reader sorts the plurality of objects of the scene based on the determined order of importance. Next, the reader indexes the objects based on the segmenting of the scene and based on the determined order of importance. Then, the reader creates a description of the scene based on the indexing, where the description is an audio, haptic, or braille representation of the scene. Next, the reader generates and conveys one or more electrical signals which include an encoding of the description of the scene to a user device to be presented on a user interface.

Abstract: A spatial audio signal is received by a spatial audio processing engine which processes the spatial audio signal to extract one or more sound features. Next, the spatial audio processing engine interprets the one or more sound features. The one or more sound features may include a direction, a distance, and an intensity of each sound source of one or more sound sources captured by the spatial audio signal. Then, the spatial audio processing engine generates textual data and/or haptic stimuli based on the interpretation of the one or more sound features. The haptic stimuli may be encoded into signals that include vibrations corresponding to a first direction and a first intensity of a first sound source. Next, the spatial audio processing engine causes the textual data and/or the haptic stimuli to be sent to a user device for presentation to a hearing-impaired user.

Abstract: A computer-implemented user-friendly system and method of designing and managing smart contracts on a distributed ledger (blockchain). The system creates a number of computer programs that correspond to a business user's model of the contract terms. In this manner, the business user can generate the smart contract without needing to understand programming or involve third parties like developers.

Abstract: The system can include a set of joints, a controller, and a model engine; and can optionally include a support structure and an end effector. Joints can include: a motor, a transmission mechanism, an input sensor, and an output sensor. The system can enable articulation of the plurality of joints.

Abstract: Disclosed herein are system, method, and computer program product embodiments for machine-assisted process modeling and validation. An embodiment operates by receiving, by at least one processor, a process document describing a process in a user locale. The embodiment then generates the model notation in accordance with a model notation format by processing the process document with a deep learning technique based on a prompt for modeling the process document. The embodiment then outputs the model notation.