Abstract: A computer-implemented method for predicting customer sentiment for a product is provided. Customer data is received for multiple products and a vector of customer sentiments is generated that is associated with different aspects of the products. Images are received for multiple products and a latent vector for the images are generated using a pre-trained image processing model. Textual description for products is received and a latent vector for the textual description is generated using a pre-trained natural language processing model. A deep multimodal design evaluation (DMDE) model is designed to integrate the latent vectors of the images and the textual descriptions to predict customer sentiment for new product designs based on their images and textual descriptions. A new product design is provided to the trained DMDE model to obtain predicted customer sentiments for one or more attributes or a new product design is generated with favorable predicted customer sentiments.

Abstract: A computer-based system and corresponding computer-implemented method for training a user via an emulated real-world building environment are provided. Based on an initial training input from a trainer user, a building simulation model is implemented. The building simulation model is configured to produce an emulated real-world building environment. For an iteration of the building simulation model, a value of at least one forcing function is determined. The at least one forcing function is associated with the emulated real-world building environment. Based on the value of the at least one forcing function determined and at least one building operation input received from a trainee user via a building training interface, a current state of the building simulation model is transformed into an updated state of the building simulation model. Based on the updated state and the initial training input, at least one building training item is generated for the trainee user.

Abstract: A biosensor that is configured to be implanted into a living being includes a light source configured to generate source light with a source light spectrum. The biosensor also includes plasmonic nano-antenna optically coupled to the light source to receive the source light and to emit light therefrom exhibiting a spectral signature of the plasmonic nano¬antenna. A bio-functionalized element joined to the plasmonic nano-antenna can receive a biomarker if available and effect a change in the spectral signature as a function of receipt or nonreceipt of the biomarker. A combination of the light source and nano-antenna can encode an optical communication signal into the emitted light and transmit the signal to a separate communication node. An optional detector can detect the emitted light and provide output indicative of receipt or nonreceipt of the biomarker, which can in turn be encoded into the communication signal for reporting to another node.

Abstract: The present technology provides devices and methods to determine the spectrum or other spectral characteristic, such as color, of a beam of light or other electromagnetic radiation. The beam of light or other electromagnetic radiation is modified without dispersion by broadband transmissive windows and then transmitted onto a detector. Signals from the detector are measured from a training set of radiation having known spectra and used to train the device, after which the device can estimate the spectrum or color of an unknown light or other electromagnetic radiation with exceptionally high accuracy.

Abstract: Provided herein are methods and systems for quasi-harmonic tags (qHT) including an electromagnetic resonator, an input mesh including an input notch filter having a resonant frequency of ?input, an output mesh including an output notch filter having a resonant frequency of ?output; and an antenna, wherein the qHT is configured to emit a comb output signal responsive to an input signal having an input frequency (op twice a resonance frequency of at least one of the electromagnetic resonators.

Abstract: Cosmetic and dermatological compositions, including color changing compositions, are provided which typically include a plurality of synthetic particles having a size in the micrometer or nanometer range. The cosmetic or dermatological formulation comprises an antioxidant or UV filter and xanthommatin pigment present in the formulation at a concentration of 0.01% wt to 0.5% wt to scavenge free radicals. The cosmetic or dermatological formulation can be a sun protectant composition having an SPF of 2-30.

Abstract: Described herein are systems, devices, and techniques that use soft-input (SI) guessing random additive noise decoding (GRAND) to provide soft-output (SO) in the form of an accurate a-posteriori estimate of the likelihood that a decoding is correct, or in the case of list decoding, the likelihood that each element of the list is correct. Also described is how to apply disclosed systems, devices, and techniques in the case where GRAND is used for decoding the component codes when the overall code is a product code or related construction, such as a generalized low-density parity-check (GLDPC). Further described is how to apply disclosed systems, devices, and techniques in the case where Polar-like decoding is used for component codes, including a framework that exploits the tradeoff between the undetected error rate (UER) and block error rate (BLER) of Polar-like codes.

Abstract: A multiple-input-multiple-output (MIMO) radar system comprises two or more software-defined millimeter-wave (SDMMW) nodes, a host processing system that is electrically coupled to the two or more SDMMW nodes, and a MIMO aperture array coupled to the two or more SDMMW nodes. The MIMO radar system is configured to form a MIMO TX channel and a MIMO RX channel for each of the two or more SDMMW nodes. The MIMO radar system includes a millimeter wave (MMW) frequency-modulated continuous wave (FMCW) oscillator source configured to generate a MMW FMCW signal. For each of the two or more SDMMW nodes, the radar system includes an upconverter assembly and a downconverter assembly.

Abstract: A true triaxial time-dependent dynamic disturbance test device for a deep-buried hard rock includes a rigid static load loading assembly and a rigid dynamic load loading assembly. The rigid static load loading assembly applies a static load to a rock test sample, and the rigid dynamic load loading assembly applies a disturbance stress to the rock test sample. Coupled application and control of a long time-dependent static force and a disturbance stress of the rock in a true triaxial disturbance stress state are achieved. The rigid dynamic load is provided with a disturbance rod in a loading direction, and a “point” disturbance mode is converted into a “face” disturbance mode through the disturbance rod, a disturbance hole and a test sample clamp, so that a disturbance mode more suitable for an actual working condition on site is provided while power consumption is reduced and time-dependent dynamic disturbance is achieved.

Abstract: Disclosed are a borehole fracture-deformation-wave velocity integrated intelligent sensing apparatus and method for an engineering rock mass. The apparatus includes an in-hole monitoring execution unit and an out-hole monitoring unit which are connected by a cable; the in-hole monitoring execution unit includes a hole-diameter-adaptive crawling robot and a combined multifunctional probe assembly; the probe assembly includes a high-definition wide-angle camera, a laser radar probe, and an acoustic transceiver probe, and all the probes may be freely combined; the hole-diameter-adaptive crawling robot is of a waterproof sealed shell structure and is capable of charging a pressure inside to realize leakage detection, and the hole-diameter-adaptive crawling robot adopts lift-type electric crawler walking mechanisms capable of automatically adjusting a supporting force to avoid sliding. The apparatus can be used for intelligent long-term monitoring of borehole fracture, deformation and wave velocity.

Abstract: A system and corresponding method sense an environment. The system comprises a wireless transmitter device that transmits a wireless signal through the environment. The system further comprises a plurality of wireless receivers that each 1) receive the wireless signal at a distinct location within the environment via at least one respective antenna and 2) generate a channel state information (CSI) packet indicating a state of a wireless communications channel associated with the wireless signal. The system still further comprises a computing device and classifier. The computing device processes the CSI packets from the plurality of wireless receivers and generates a CSI dataset as a function of the CSI packets processed. The classifier determines at least one class for the CSI dataset. The system and corresponding method improve robustness of sensing operations, such as robustness of Wi-Fi sensing operations to noise and interference.

Abstract: An ambulatory glucose profile (AGP) intelligent interpretation and insulin adjustment method based on an expert system includes: establishing a knowledge base in an inference mechanism; constructing an interpretation and decision support expert system with a simplified expert system architecture based on the knowledge base; constructing a patient problem analysis tree in three dimensions of hypoglycemia, blood glucose fluctuation, and hyperglycemia of patients; expanding each rule with expert AGP interpretation and empirical data; constructing a basal insulin dosage adjustment rule and a mealtime insulin dosage adjustment rule based on an interval type-2 fuzzy expert system; and adjusting a node of the patient problem analysis tree based on the interpretation and decision support expert system and a group of the patient, and providing a decision suggestion in combination with the basal insulin dosage adjustment rule and the mealtime insulin dosage adjustment rule.

Abstract: A power conversion device includes an input stage from an unregulated power source to a switching circuit that is coupled to a thin film link capacitor. In some embodiments, the link stage permits soft switching. Further, a high frequency transformer may be utilized for management of voltage acquired from the circuit. Thus, a managed voltage may be supplied to an electrical load via a capacitor of reduced size and weight and with improved efficiency and reliability.

Abstract: The present invention provides a compact steel plate continuous quenching machine, which includes a load-bearing steel frame, a support transmission assembly, a water supply mechanism, a first water supply pipeline assembly, a first nozzle assembly, a support steel frame, a lifting mechanism, a movable steel frame, a compression transmission assembly, a second water supply pipeline assembly, a water jacket, a third water supply pipeline assembly, and a second nozzle assembly. By enabling the third water supply pipeline assembly to move within the water jacket to adapt to the height of the movable steel frame, compared to using a metal hose to connect the second nozzle assembly with the second water supply pipeline assembly, the space occupied by the bending of the metal hose can be saved, thereby making the product more compact, reducing the space occupied by the product, and further enhancing the user experience of the product.

Abstract: Me-N—C catalysts, wherein Me can include a transition metal, Mn, Fe, Co, or a combination of metals with Me-INU moieties located at the exterior surface of the Me-N—C catalysts are produced by a chemical vapor deposition synthesis. The synthesis methods can utilize non-solid-contact pyrolysis wherein a metal salt can be vaporized. Gaseous metal from the vaporized metal salt can displace a metal M from the N—C zeolitic imidazolate framework. The non-solid-contact pyrolysis does not mix solid iron precursors (e.g., Me?Mn, Fe, or Co) with the solid N—C zeolitic imidazolate framework precursors during or before the synthesis, which improves the process compared to conventional methods.

Abstract: A method for broadband, high-efficiency spin wave transduction adopts a slow-wave structure to enhance the interaction of electromagnetic waves and spin waves.

Abstract: A machine learning system and method for optimizing a production process. For instance, the method includes several steps as follows: selecting different values for a plurality of input parameters of a digital model of the production process for simulation; running the digital model using the different values for the plurality of input parameters and at least some of real-time data of the production process; determining a plurality of output parameters of the digital model; analyzing the plurality of output parameters; learning an optimized plurality of input parameters corresponding to the plurality of output parameters; and programming the production process to use the optimized plurality of input parameters to run the production process.

Abstract: Described herein is a sensor comprising a doped silicon layer, a graphene layer on the doped silicon layer, a molecularly imprinted polymer (MIP) layer on the graphene layer, and electrodes in operative arrangement with the MIP layer and configured to provide a signal indicative of resistance. The MIP layer is derived from a MIP monomer and functional monomer. Also described herein is a detector comprising a sensor described herein as well as methods of making and using the sensors and detectors, e.g., to detect an analyte, such as a virus.

Abstract: Aspects of the present disclosure describe systems and methods for predicting an intra-aortic pressure of a patient receiving hemodynamic support from a transvalvular micro-axial heart pump. In some implementations, an intra-aortic pressure time series is derived from measurements of a pressure sensor of the transvalvular micro-axial heart pump and a motor speed time series is derived from a measured back electromotive force of a motor of the transvalvular micro-axial heart pump. Furthermore, in some implementations, machine learning algorithms, such as deep learning, are applied to the intra-aortic pressure and motor speed time series to accurately predict an intra-aortic pressure of the patient. In some implementations, the prediction is short-term (e.g., approximately 5 minutes in advance).

Abstract: ADMM-NN is an algorithm-hardware co-optimization framework of DNNs using Alternating Direction Method of Multipliers (ADMM). The first part of ADMM-NN is a systematic, joint framework of DNN weight pruning and quantization using ADMM. The second part is a hardware-aware optimization to facilitate hardware-level implementations. ADMM-based weight pruning and quantization accounts for (i) computation reduction and energy efficiency improvement and (ii) performance overhead due to irregular sparsity. Experimental results demonstrate that by combining weight pruning and quantization, the proposed framework can achieve 1,910× and 231× reductions in the overall model size on the LeNet-5 and AlexNet models. Favorable results are also observed on VGGNet and ResNet models. Also, without any accuracy loss, 3.6× reduction in the amount of computation can be achieved, outperforming prior work.