Abstract: An embodiment includes training a first convolutional neural network (CNN) using a plurality of training images to generate first and second trained CNNs, and then adding an interface layer to the second trained CNN. The embodiment processes a first and second images in a sequence of images using the first trained CNN to generate a first and second result vectors. The embodiment also processes the second image using the second trained CNN and sensor data input to the interface layer to generate a third result vector. The embodiment modifies the sensor data using a compensation value. The embodiment compares the third result vector to the second result vector to generate an error value, and then calculates a modified compensation value using the error value. The embodiment then generates a sensor-compensated trained CNN based on the second trained CNN with the modified compensation value.

Abstract: A pretrained model is selected to operate in an augmented model configuration with a submodel. The submodel is trained using training data corresponding to a second domain, whereas the pretrained model is trained to operate on data of a first domain. The pretrained model is augmented, to form the augmented model configuration, with the submodel, by combining a first feature map being output from a layer in the pretrained model with a second feature map being output from a layer in the submodel. The combining forms a combined feature map. The combined feature map is input into a different layer in the submodel.

Abstract: A radio frequency (RF) drying device is provided. The RF drying device includes a drying chamber including a grounded body enclosing a plurality of N (N?3) electrodes for generating radio frequency field, and a drum electrically insulated from the plurality of electrodes, and a processor configured to select and activate M (<N) electrodes among the plurality of N electrodes and deactivate one or more remaining (N?M) electrodes periodically and to to select M(<N) electrodes for being activated to generate AC signals and deactivate N?M electrodes for being inactivated not to generate AC signals in an active mode, and RF field is formed inside the drying chamber based on the generated AC signals.

Abstract: An electronic device may comprise: a load circuit including a plurality of electrodes; a converter comprising circuitry configured to convert the magnitude of direct current power; at least one driving circuit configured to output alternating current power to the load circuit using DC power provided from the converter; a temperature sensor configured to measure the temperature of an object arranged between the plurality of electrodes; and at least one controller comprising circuitry.

Abstract: Provided is an electronic device configured to accommodate at least one container, the electronic device including a conductive plate configured to form an electric field based on a container being accommodated at a designated position of the electronic device, and a conductive member connected to the conductive plate and having an adjustable length, the conductive member being configured to adjust a position of the conductive plate based on the adjustable length, wherein the conductive plate includes a plurality of conductive plates that are spaced apart from each other by a gap formed along a direction parallel to a longitudinal cross-section of the container, and wherein, based on the container being accommodated at the designated position, the plurality of conductive plates are configured to form an electric field in a transverse cross-section direction of the container.

Abstract: A dielectric heating device is provided. The dielectric heating device includes at least one power source, a load circuit including a plurality of electrodes and at least one inductor, and at least one driving circuit configured to output alternating-current power to the load circuit by using power provided from the at least one power source, wherein the plurality of electrodes include a plurality of top electrodes arranged in substantially the same plane, and one bottom electrode arranged in a plane parallel to the plane formed by the plurality of top electrodes.

Abstract: The subject technology provides memory-efficient differentiable weight clustering for large language model compression. An apparatus determines a tensor including an attention map between learned weights of a trained machine learning model and corresponding centroids. The apparatus also determines a compressed attention table and a plurality of index lists during compression of the trained machine learning model based on an uniquification of the attention map and sharding of an associated index list. The apparatus determines whether the tensor exists at a destination device during compression of the trained machine learning model using a marshaling layer. The apparatus refrains from copying the tensor to the destination device when the tensor exists at the destination device, or copies the tensor to the destination device when the tensor does not exist at the destination device. The apparatus deploys a compressed machine learning model based on the compression of the trained machine learning model.

Abstract: A batched Cholesky decomposition method, system, and non-transitory computer readable medium for a Graphics Processing Unit (GPU), include mirroring matrices to form paired matrices solving the paired matrices simultaneously.

Abstract: An overall gradient vector is computed at a server from a set of ISA vectors corresponding to a set of worker machines. An ISA vector of a worker machine including ISA instructions corresponding to a set of gradients, each gradient corresponding to a weight of a node of a neural network being distributedly trained in the worker machine. A set of register values is optimized for use in an approximation computation with an opcode to produce an x-th approximate gradient of an x-th gradient. A server ISA vector is constructed in which a server ISA instruction in an x-th position corresponds to the x-th gradient in the overall gradient vector. A processor at the worker machine is caused to update a set of weights of the neural network, using the set of optimized register values and the server ISA vector, thereby completing one iteration of training.

Abstract: Techniques that facilitate model support in deep learning are provided. In one example, a system includes a graphics processing unit and a central processing unit memory. The graphics processing unit processes data to train a deep neural network. The central processing unit memory stores a portion of the data to train the deep neural network. The graphics processing unit provides, during a forward pass process of the deep neural network that traverses through a set of layers for the deep neural network from a first layer of the set of layers to a last layer of the set of layers that provides a set of outputs for the deep neural network, input data for a layer from the set of layers for the deep neural network to the central processing unit memory.

Abstract: A computer-implemented method comprises linking a private AI model to a public AI model to thereby form a combined AI model comprising the private AI model and the public AI model; and training the combined AI model with private samples while keeping the public AI model fixed so that only the private AI model is trained with the private samples.

Abstract: A batched Cholesky decomposition method, system, and non-transitory computer readable medium for a Graphics Processing Unit (GPU), include mirroring matrices to form paired matrices solving the paired matrices simultaneously.

Abstract: A batched Cholesky decomposition method, system, and non-transitory computer readable medium for a Graphics Processing Unit (GPU), include mirroring matrices to form paired matrices solving the paired matrices simultaneously.

Abstract: Techniques that facilitate model support in deep learning are provided. In one example, a system includes a graphics processing unit and a central processing unit memory. The graphics processing unit processes data to train a deep neural network. The central processing unit memory stores a portion of the data to train the deep neural network. The graphics processing unit provides, during a forward pass process of the deep neural network that traverses through a set of layers for the deep neural network from a first layer of the set of layers to a last layer of the set of layers that provides a set of outputs for the deep neural network, input data for a layer from the set of layers for the deep neural network to the central processing unit memory.

Abstract: An embodiment includes training a first convolutional neural network (CNN) using a plurality of training images to generate first and second trained CNNs, and then adding an interface layer to the second trained CNN. The embodiment processes a first and second images in a sequence of images using the first trained CNN to generate a first and second result vectors. The embodiment also processes the second image using the second trained CNN and sensor data input to the interface layer to generate a third result vector. The embodiment modifies the sensor data using a compensation value. The embodiment compares the third result vector to the second result vector to generate an error value, and then calculates a modified compensation value using the error value. The embodiment then generates a sensor-compensated trained CNN based on the second trained CNN with the modified compensation value.

Abstract: Techniques that facilitate model support in deep learning are provided. In one example, a system includes a graphics processing unit and a central processing unit memory. The graphics processing unit processes data to train a deep neural network. The central processing unit memory stores a portion of the data to train the deep neural network. The graphics processing unit provides, during a forward pass process of the deep neural network that traverses through a set of layers for the deep neural network from a first layer of the set of layers to a last layer of the set of layers that provides a set of outputs for the deep neural network, input data for a layer from the set of layers for the deep neural network to the central processing unit memory.

Abstract: Various embodiments are provided for decentralized distributed deep learning by one or more processors in a computing system. Asynchronous distributed training of one or more machine learning models may be performed by generating a list of neighbor nodes for each node in a plurality of nodes and creating a first thread for continuous communication according to a weight management operation and a second thread for continuous computation of a gradient for each node. One or more variables are shared between the first thread and the second thread.

Abstract: In deep learning, and in particular, for data compression for allreduce in deep learning, a gradient may be compressed for synchronization in a data parallel deep neural network training for allreduce by sharing a consensus vector between each node in a plurality of nodes to ensure identical indexing in each of the plurality of nodes prior to performing sparse encoding.

Abstract: A computer-implemented method generates a hamming code based target label for each class of a dataset in which hamming distance between the target labels in the dataset is maximized and trains a convolutional neural network with the hamming codes based target label to thereby produce a trained AI model. The confusability between classes of the dataset is determined using a confusion matrix. The hamming distances of classes of the dataset that are determined to be more confusable are set to higher values than the hamming distances of classes of the dataset that are determined to be less confusable.

Abstract: A computer-implemented method comprises: inputting into an autoencoder sets of input samples, each of the sets of input samples comprising: a reference input sample of a reference dataset and one or more target input samples of one or more target datasets, the autoencoder being trained using the reference dataset. The autoencoder generates a respective set of outputs for each set of the input samples to thereby form one or more respective sets of outputs, each of the one or more sets of outputs comprising the reference output and the one or more target outputs for a respective set of input samples; and determining the similarity of each of the one or more target datasets to the reference dataset by comparing each of the one or more target outputs to respective target input samples of each of the sets of input samples.