Abstract: Provided are computing systems, methods, and platforms that rank text with pairwise ranking prompting using a generative sequence processing model. A prompt comprising a query and sets of text associated with candidate results can be generated. The generative sequence processing model can be prompted with the prompt and perform pairwise comparisons between the sets of text in the prompt based on the query in the prompt. An output can be generated that ranks the sets of text in response to the query.

Abstract: Smile treatment planning systems and methods are described herein. One method for adjusting an image of a smile may generally comprise receiving a three-dimensional (3D) digital model of a dental arch of a patient, receiving a digital facial image of the patient which includes an image of one or more teeth of the patient when smiling, registering the 3D digital model to the one or more teeth of the patient from the digital facial image, correcting the 3D digital model for scale and distortion to create a corrected 3D digital model, and overlaying the corrected 3D digital model onto the digital facial image.

Abstract: A neural architecture search (NAS) with a weak predictor comprises: receiving network architecture scoring information; iteratively sampling a search space, wherein the sampling comprises: generating a set of candidate architectures within the search space; learning a first predictor; evaluating performance of the candidate architectures; and based on at least the performance of the set of candidate architectures and the network architecture scoring information, refining the search space to a smaller search space; based on at least the network architecture scoring information, thresholding the performance of candidate architectures to determine scored output candidate architectures; and reporting the scored output candidate architectures. In some examples, the candidate architectures each comprise a machine learning (ML) model, for example a neural network (NN).

Abstract: A computer-implemented method for implementing face spoofing detection using a physical-cue-guided multi-source multi-channel framework includes receiving a set of data including face recognition data, liveness data and material data associated with at least one face image, obtaining a shared feature from the set of data using a backbone neural network structure, performing, based on the shared feature, a pretext task corresponding to face recognition, a first proxy task corresponding to depth estimation, a liveness detection task, and a second proxy task corresponding to material prediction, and aggregating outputs of the pretext task, the first proxy task, the liveness detection task and the second proxy task using an attention mechanism to boost face spoofing detection performance.

Abstract: A computer-implemented method for implementing face spoofing detection using a physical-cue-guided multi-source multi-channel framework includes receiving a set of data including face recognition data, liveness data and material data associated with at least one face image, obtaining a shared feature from the set of data using a backbone neural network structure, performing, based on the shared feature, a pretext task corresponding to face recognition, a first proxy task corresponding to depth estimation, a liveness detection task, and a second proxy task corresponding to material prediction, and aggregating outputs of the pretext task, the first proxy task, the liveness detection task and the second proxy task using an attention mechanism to boost face spoofing detection performance.

Abstract: The present invention relates to an energy harvester system. The energy harvester system includes an energy harvester device, a housing comprising internal walls surrounding at least a portion of the energy harvester device, and a flexible supporting structure supporting the energy harvester device within the housing. Movement of the housing causes the internal walls of the housing, or structures connected to the internal walls, to contact the energy harvester device, whereby the flexible supporting structure and the energy harvester device move such that the energy harvester device contacts the internal walls or structures connected to the internal walls at least one additional time (or multiple times), thereby producing energy. Also disclosed is a system comprising an electrically powered apparatus and the energy harvester system of the present invention electrically coupled to the apparatus.

Abstract: The aero-acoustic duster invention disclosed herein provides for high particle removal rate from surfaces with low energy expenditure relative to competing vacuum-based devices. The device removes particulate matter from a surface using a two-step process: 1. Acoustic radiation is used to break the adhesive bonds between dust and the surface, forcing particles into a mode where they continuously bounce up and down on the surface; and, 2. A bounded vortex is generated over the surface, with suction in the vortex center and jets for blowing air along the periphery. The jets are tilted in the tangential direction to induce vortex motion within the suction region. The vortex is said to be bounded because streamlines originating in the downward jets are entrained back into the central vortex.

Abstract: Vibrational energy harvesting (VEH) structures that include resonant beams each having a fundamental resonance frequency and a parametric mode frequency and including at least one piezoelectric layer for generating electrical charge in response to each of fundamental-resonance excitation and parametric-mode excitation of that beam. Circuitry is provided for harvesting the electrical charge from the resonant beam. In some embodiments, the parametric mode frequency of the beam is tuned to be close to its fundamental resonance frequency so as to increase the effective bandwidth of a VEH structure. The effective bandwidth of a VEH structure can be further increased by tuning ones of multiple parametric-mode-enabled resonant beams to slightly different fundamental resonance frequencies and parametric mode frequencies.

Abstract: The aero-acoustic duster invention disclosed herein provides for high particle removal rate from surfaces with low energy expenditure relative to competing vacuum-based devices. The device removes particulate matter from a surface using a two-step process: 1. Acoustic radiation is used to break the adhesive bonds between dust and the surface, forcing particles into a mode where they continuously bounce up and down on the surface; and, 2. A bounded vortex is generated over the surface, with suction in the vortex center and jets for blowing air along the periphery. The jets are tilted in the tangential direction to induce vortex motion within the suction region. The vortex is said to be bounded because streamlines originating in the downward jets are entrained back into the central vortex.

Abstract: Vibrational energy harvesting (VEH) structures that include resonant beams each having a fundamental resonance frequency and a parametric mode frequency and including at least one piezoelectric layer for generating electrical charge in response to each of fundamental-resonance excitation and parametric-mode excitation of that beam. Circuitry is provided for harvesting the electrical charge from the resonant beam. In some embodiments, the parametric mode frequency of the beam is tuned to be close to its fundamental resonance frequency so as to increase the effective bandwidth of a VEH structure. The effective bandwidth of a VEH structure can be further increased by tuning ones of multiple parametric-mode-enabled resonant beams to slightly different fundamental resonance frequencies and parametric mode frequencies.

Abstract: A system for detecting wear debris particulate from a medical implant within the body of a living animal is provided. The system includes an acoustic transmitter for transmitting acoustic energy from outside the body to a soft tissue region proximate the medical implant containing wear debris particles; an acoustic receiver located outside the body to detect resultant acoustic energy generated by the wear debris particles and produce a received signal indicative thereof; a processor for processing the received signal to evaluate at least one parameter associated with the wear debris particles; and an output for indicating the at least one parameter.