Abstract: An adapter to a base model of an artificial intelligence (AI) system is disclosed. The adapter includes a connector to connect the adapter to the base model such that during an operation of the AI system at least some portion of data transformed by the base model is propagated from the base model to the adapter and back from the adapter to the base model. The adapter includes a non-linear modifier to modify the data received from the base model non-linearly before returning the modified portion of the data back to the base model, and an AI trainer to tune the non-linear modifier of the adapter by propagating training data through the base model and the adapter and updating weights of the non-linear modifier of the adapter for given weights of the base model to optimize a loss function. Further, weight matrices for the base model and the adapter are jointly constructed by an additional module, which efficiently uses a pool of parameters to allocate to save memory requirement for adaptation of the AI system.

Abstract: Techniques for code-free automated machine learning (ML) are described. Users can train high-quality ML models and pipelines without necessarily needing to write code by providing a training dataset to a code-free machine learning service. The service may deploy an ML orchestration function and a storage location on behalf of a user. When a modification is made to the storage bucket, such as by the user providing a training dataset, the orchestration function is invoked and can automatically initiate an AutoML process using at least the training data to train multiple ML model variants. The resultant ML model(s) and associated metrics can be provided to the user, deployed behind an endpoint, and/or used to generate inferences.

Abstract: A robust machine learning system is provided for robust machine learning. The robust machine learning system includes a processor, and a memory storing a robust machine learning program and a Group Representation Network (GRepsNet), wherein the robust machine learning program runining on the processor. In this case the processor is configured to receive an input dataset represented by a first group representation via input layers connected to the GrepsNet, pass/transmit the input dataset through the GRepsNet configured to generate an output dataset represented by an output group representation from the input dataset, wherein the output group is equivalent to the first group, and generate the output dataset from output layers connected to the GRepsNet.

Abstract: A camera is provided for reconstructing the surface of the object. The camera includes an optical module configured to spatially separate photons reflected from an object in a scene to form at least two focused and distinctly polarized images on a sensor, wherein the sensor is configured to receive the at least two polarized images on pixels the sensor and generate intensity values of the pixels, and a computing module including a processor and a memory having instructions stored thereon. According to the instructions, the processor computes depth values at points of a surface of the object based on a ratio of the intensity values of the pixels with respect to the at least two polarized images and reconstructs the surface of the object from the computed depth values.

Abstract: A method for determining a spatial distribution of a radiation damage of a heterogenous material. The steps include some or all of the following. Representing a multiphase microstructure of the heterogenous material as a 2D or 3D image. Determine a first energy of a primary knock-on atom (PKA) ion at a first interface at a first distance from an incident PKA ion to a first phase material of the microstructure image representation. Then determine a second PKA energy in a second phase across a first interface using the first PKA energy and PKA energy-depth-damage profiles of the same PKA in the bulk (isolated) materials of the parent phases. The process is repeated at subsequent interfaces. The radiation damage being determined from the PKA ion energy deposited in each phase material.

Abstract: A camera is provided for reconstructing the surface of the object. The camera includes an optical module configured to spatially separate photons reflected from an object in a scene to form at least two focused and distinctly polarized images on a sensor, wherein the sensor is configured to receive the at least two polarized images on pixels the sensor and generate intensity values of the pixels, and a computing module including a processor and a memory having instructions stored thereon. According to the instructions, the processor computes depth values at points of a surface of the object based on a ratio of the intensity values of the pixels with respect to the at least two polarized images and reconstructs the surface of the object from the computed depth values.

Abstract: A reconfigurable device is provided for splitting optical beams. The device includes an input port configured to receive an input beam including at least two primary wavelengths, a tunable splitter configured to separate the input beam into at least two beams via at least two routes corresponding to the at least two primary wavelengths, wherein each of the at least two routes is configured to propagate one of the at least two primary wavelengths of the input beam, wherein the tunable splitter includes a bottom electrode, a substrate on the bottom electrode, core segments arranged on the substrate, a top layer, support segments to connect the substrate and the top layer, a top electrode on the top layer, a controllable refractive index layer arranged to fill gaps between the substrate, the support segments, and the top layer; and at least two output ports configured to transmit the at least two beams propagated via the at least two routes.

Abstract: A computer-implemented method is provided for designing a metalens. The metalens has a metasurface including nano-structures arranged on a substrate based on preliminary simulation data.

Abstract: Embodiments of the present disclosure disclose a method and a system for training a neural network for improving adversarial robustness. The method includes collecting a plurality of data samples comprising clean data samples and adversarial data samples. The training of the neural network includes training of a probabilistic encoder to encode the plurality of data samples into a probabilistic distribution over a latent space representation. In addition, the training of the neural network comprising training of a classifier to classify an instance of the latent space representation to produce a classification result. In addition, the method includes training shared parameters of a first instance of the neural network using the clean data samples and a second instance of the neural network using the adversarial data samples. Further, the method includes outputting the shared parameters of the first instance of the neural network and the second instance of the neural network.

Abstract: A system for providing resistance in an exercise machine. The system includes a motor, at least one loop drive attached to the motor. A carriage is coupled to the loop drive. The carriage moves in a first direction when the motor is turned in a first direction and a second direction when the motor is turned in a second direction. At least one sensor is attached to the carriage, wherein the at least one sensor is configured to detect external force on the carriage. Information from the sensor indicates external force on the carriage. The information is used to determine a movement of the carriage in response to the external force. The motor is instructed to turn the loop drive to apply the movement.

Abstract: A data conversion system includes an interface to receive path data, a memory to store a computer-executable program including a lattice full algorithm and a dynamic programming algorithm, a processor, in connection with the memory, configured to execute the computer-executable program.

Abstract: A computer-based system trains a neural network by solving a double layer optimization problem. The system includes an input interface to receive an input to the neural network and labels of the input to the neural network; a processor to solve a double layer optimization to produce parameters of the neural network, and an output interface to output the parameters of the neural network. The double layer optimization includes an optimization of a first layer subject to an optimization of a second layer. The optimization of the first layer minimizes a difference between an output of the neural network processing the input and the labels of the input to the neural network, the optimization of the second layer minimizes a distance between a non-negative output vector of each layer and a corresponding input vector to each layer. The input vector of a current layer is a linear transformation of the non-negative output vector of the previous layer.

Abstract: Techniques for code-free automated machine learning (ML) are described. Users can train high-quality ML models and pipelines without necessarily needing to write code by providing a training dataset to a code-free machine learning service. The service may deploy an ML orchestration function and a storage location on behalf of a user. When a modification is made to the storage bucket, such as by the user providing a training dataset, the orchestration function is invoked and can automatically initiate an AutoML process using at least the training data to train multiple ML model variants. The resultant ML model(s) and associated metrics can be provided to the user, deployed behind an endpoint, and/or used to generate inferences.

Abstract: A lighting apparatus for providing uniform irradiance is provided. The lighting apparatus includes a light source configured to emit rays toward a polygonal target region, and a lens configured to uniformize irradiance from the light source to the target region, wherein sag of the lens and normals along a boundary of the lens are arranged so as to illuminate a perimeter of the target region based on an edge-ray mapping, wherein a lens surface is formed based on polyharmonic spline interpolations so as to smoothly illuminate an interior of the target region, wherein the irradiance is uniformized by argumenting light field tailoring of the lens, wherein the light field tailoring is augmented to maintain to maintain the irradiance perimeter.

Abstract: A system for providing resistance in an exercise machine. The system includes a motor, at least one loop drive attached to the motor. A carriage is coupled to the loop drive. The carriage moves in a first direction when the motor is turned in a first direction and a second direction when the motor is turned in a second direction. At least one sensor is attached to the carriage, wherein the at least one sensor is configured to detect external force on the carriage. Information from the sensor indicates external force on the carriage. The information is used to determine a movement of the carriage in response to the external force. The motor is instructed to turn the loop drive to apply the movement.

Abstract: An exemplary lockset includes a first bolt, a second bolt, and a drive assembly operable to retract the first bolt and the second bolt. The drive assembly includes a first gear train including a first input gear and a first output gear operably connected with the first bolt, a second gear train including a second input gear and a second output gear operably connected with the second bolt, a third input gear, and an electromechanical driver operable to rotate the third input gear in an unlocking direction. The third input gear is engaged with the first input gear and the second input gear such that rotation of the third input gear in the unlocking direction causes rotation of the first and second output gears, thereby retracting the first and second bolts.

Abstract: Devices, systems and methods including an optic having an optical surface partitioned into a plurality of facets. Wherein a geometry of the plurality of facets are configured to rearrange the rays in an incident light field to form a desired irradiance pattern on a projection surface, partitioned into a plurality of pixels, each of which may receive light from a multiplicity of facets. Wherein the configured geometry of the plurality of facets is corrected to minimize a difference between an observed irradiance pattern and a desired irradiance pattern provided by a user.

Abstract: A data conversion system includes an interface to receive path data, a memory to store a computer-executable program including a lattice full algorithm and a dynamic programming algorithm, a processor, in connection with the memory, configured to execute the computer-executable program.

Abstract: Devices, systems and methods including an optic having an optical surface partitioned into a plurality of facets. Wherein a geometry of the plurality of facets are configured to rearrange the rays in an incident light field to form a desired irradiance pattern on a projection surface, partitioned into a plurality of pixels, each of which may receive light from a multiplicity of facets. Wherein the configured geometry of the plurality of facets is corrected to minimize a difference between an observed irradiance pattern and a desired irradiance pattern provided by a user.

Abstract: A system for providing resistance in an exercise machine. The system includes a motor, at least one loop drive attached to the motor. A carriage is coupled to the loop drive. The carriage moves in a first direction when the motor is turned in a first direction and a second direction when the motor is turned in a second direction. At least one sensor is attached to the carriage, wherein the at least one sensor is configured to detect external force on the carriage. Information from the sensor indicates external force on the carriage. The information is used to determine a movement of the carriage in response to the external force. The motor is instructed to turn the loop drive to apply the movement.