Abstract: A system for image quality assessment of non-aligned images includes a first deep path portion of a convolutional neural network having a set of parameters and a second deep path portion of the convolutional neural network sharing a set of parameters with the first deep path convolutional neural network. Weights are shared between the first and second deep path convolutional neural networks to support extraction of a same set of features in each neural network pathway. Non-aligned reference and distorted images are respectively provided to the first and second deep paths of the convolutional neural network for processing. A concatenation layer is connected to both the first and second deep paths convolutional neural network, and a fully connected layer is connected to the concatenation layer to receive input from both the first and second deep paths of the convolutional neural network, generating an image quality assessment as a linear regressor and outputting an image quality score.

Abstract: Self-checkout verification systems and methods are described. One aspect includes receiving a plurality of images from a camera, the images being associated with a customer self-checkout process. The images may be analyzed to detect one or more items in possession of the customer, count a first number of the items, categorize each item, and construct a first category set including the categorizing for all the items. One aspect includes receiving a point-of-sale record at a completion of the self-checkout process, the point-of-sale record including a second number of the items and a second category set including the categorizing for all the items. The first number and the second number, and the first category set and the second category set may be compared. An alert may be generated if there is a discrepancy between the first number and the second number, or the first category set and the second category set.

Abstract: A semi-supervised model incorporates deep feature learning and pseudo label estimation into a unified framework. The deep feature learning can include multiple convolutional neural networks (CNNs). The CNNs can be trained on available training datasets, tuned using a small amount of labeled training samples, and stored as the original models. Features are then extracted for unlabeled training samples by utilizing the original models. Multi-view clustering is used to cluster features to generate pseudo labels. Then the original models are tuned by using an updated training set that includes labeled training samples and unlabeled training samples with pseudo labels. Iterations of multi-view clustering and tuning using an updated training set can continue until the updated training set is stable.

Abstract: The drilling system withstands downhole conditions at the bottom of the borehole and drilling conditions due to the constant movement and vibration. The drilling system includes a drill bit and a sensor system with a system housing, a primary power supply and interior sensor. The primary power supply includes piezoelectric panels for converting radial vibration into energy. The interior sensor is locally powered by the primary power supply at the remote downhole location at the bottom of the borehole. The interior sensor collects data related to a downhole condition and is in communication with the primary power supply to generate confirmed data based on the amount of energy generated by the primary power supply. The confirmed data is more accurate and reliable than the data collected by the interior sensor and can be used to guide the path of the drill bit through the rock formation in drilling operations.

Abstract: The drilling system withstands downhole conditions at the bottom of the borehole and drilling conditions due to the constant movement and vibration. The drilling system includes a drill bit and a sensor system with a system housing, a primary power supply and interior sensor. The primary power supply includes piezoelectric panels for converting radial vibration into energy. The interior sensor is locally powered by the primary power supply at the remote downhole location at the bottom of the borehole. The interior sensor collects data related to a downhole condition and is in communication with the primary power supply to generate confirmed data based on the amount of energy generated by the primary power supply. The confirmed data is more accurate and reliable than the data collected by the interior sensor and can be used to guide the path of the drill bit through the rock formation in drilling operations.

Abstract: Provided are a method and apparatus for interaction in a three-dimensional space. The method includes: displaying a target augmented reality three-dimensional picture on a target augmented reality device, and displaying a target virtual interaction object in a virtual three-dimensional space in the target augmented reality three-dimensional picture, wherein the real three-dimensional space is rasterized and the virtual three-dimensional space is rasterized, and there is a mapping relationship therebetween; acquiring a target movement trajectory of a target object in the real three-dimensional space; determining a real cell set through which the target movement trajectory passes; determining a passing first virtual cell set corresponding to the passing real cell set; and determining whether the target movement trajectory triggers a target interaction operation of the target virtual interaction object according to the position relationship between the first virtual cell set and a second virtual cell set.

Abstract: The present disclosure provides a device and method for calibrating a particle image velocimetry (PIV) image based on laser linear arrays, and relates to the technical field of laser velocity measurement and image restoration. The present disclosure can solve the problem of image distortion caused by a shock wave of a model in a hypersonic wind tunnel, thereby realizing distortion capture and correction. The device includes a laser emission component configured to emit equidistant laser linear arrays; an optical component configured to perform light splitting on laser rays to form a laser grating in a test observation region; a camera configured to acquire a distorted laser grating image when a working condition of a wind tunnel test section model is adjusted to a working condition of a PIV test; and a background processor configured to calibrate and restore the distorted laser grating image with a neural network-based distortion-restoring calibration algorithm.

Abstract: The present disclosure provides a velocity measurement method, system, device and apparatus and storage medium, as well as a velocity field measurement method and system.

Abstract: A system for image quality assessment of non-aligned images includes a first deep path portion of a convolutional neural network having a set of parameters and a second deep path portion of the convolutional neural network sharing a set of parameters with the first deep path convolutional neural network. Weights are shared between the first and second deep path convolutional neural networks to support extraction of a same set of features in each neural network pathway. Non-aligned reference and distorted images are respectively provided to the first and second deep paths of the convolutional neural network for processing. A concatenation layer is connected to both the first and second deep paths convolutional neural network, and a fully connected layer is connected to the concatenation layer to receive input from both the first and second deep paths of the convolutional neural network, generating an image quality assessment as a linear regressor and outputting an image quality score.

Abstract: A semi-supervised model incorporates deep feature learning and pseudo label estimation into a unified framework. The deep feature learning can include multiple convolutional neural networks (CNNs). The CNNs can be trained on available training datasets, tuned using a small amount of labeled training samples, and stored as the original models. Features are then extracted for unlabeled training samples by utilizing the original models. Multi-view clustering is used to cluster features to generate pseudo labels. Then the original models are tuned by using an updated training set that includes labeled training samples and unlabeled training samples with pseudo labels. Iterations of multi-view clustering and tuning using an updated training set can continue until the updated training set is stable.

Abstract: A facial recognition method using online sparse learning includes initializing target position and scale, extracting positive and negative samples, and extracting high-dimensional Haar-like features. A sparse coding function can be used to determine sparse Haar-like features and form a sparse feature matrix, and the sparse feature matrix in turn is used to classify targets.

Abstract: A semi-supervised model incorporates deep feature learning and pseudo label estimation into a unified framework. The deep feature learning can include multiple convolutional neural networks (CNNs). The CNNs can be trained on available training datasets, tuned using a small amount of labeled training samples, and stored as the original models. Features are then extracted for unlabeled training samples by utilizing the original models. Multi-view clustering is used to cluster features to generate pseudo labels. Then the original models are tuned by using an updated training set that includes labeled training samples and unlabeled training samples with pseudo labels. Iterations of multi-view clustering and tuning using an updated training set can continue until the updated training set is stable.

Abstract: Self-checkout verification systems and methods are described. One aspect includes receiving a plurality of images from a camera, the images being associated with a customer self-checkout process. The images may be analyzed to detect one or more items in possession of the customer, count a first number of the items, categorize each item, and construct a first category set including the categorizing for all the items. One aspect includes receiving a point-of-sale record at a completion of the self-checkout process, the point-of-sale record including a second number of the items and a second category set including the categorizing for all the items. The first number and the second number, and the first category set and the second category set may be compared. An alert may be generated if there is a discrepancy between the first number and the second number, or the first category set and the second category set.

Abstract: A system for image quality assessment of non-aligned images includes a first deep path portion of a convolutional neural network having a set of parameters and a second deep path portion of the convolutional neural network sharing a set of parameters with the first deep path convolutional neural network. Weights are shared between the first and second deep path convolutional neural networks to support extraction of a same set of features in each neural network pathway. Non-aligned reference and distorted images are respectively provided to the first and second deep paths of the convolutional neural network for processing. A concatenation layer is connected to both the first and second deep paths convolutional neural network, and a fully connected layer is connected to the concatenation layer to receive input from both the first and second deep paths of the convolutional neural network, generating an image quality assessment as a linear regressor and outputting an image quality score.

Abstract: A semi-supervised model incorporates deep feature learning and pseudo label estimation into a unified framework. The deep feature learning can include multiple convolutional neural networks (CNNs). The CNNs can be trained on available training datasets, tuned using a small amount of labeled training samples, and stored as the original models. Features are then extracted for unlabeled training samples by utilizing the original models. Multi-view clustering is used to cluster features to generate pseudo labels. Then the original models are tuned by using an updated training set that includes labeled training samples and unlabeled training samples with pseudo labels. Iterations of multi-view clustering and tuning using an updated training set can continue until the updated training set is stable.

Abstract: A facial recognition method using online sparse learning includes initializing target position and scale, extracting positive and negative samples, and extracting high-dimensional Haar-like features. A sparse coding function can be used to determine sparse Haar-like features and form a sparse feature matrix, and the sparse feature matrix in turn is used to classify targets.

Abstract: A semi-supervised model incorporates deep feature learning and pseudo label estimation into a unified framework. The deep feature learning can include multiple convolutional neural networks (CNNs). The CNNs can be trained on available training datasets, tuned using a small amount of labeled training samples, and stored as the original models. Features are then extracted for unlabeled training samples by utilizing the original models. Multi-view clustering is used to cluster features to generate pseudo labels. Then the original models are tuned by using an updated training set that includes labeled training samples and unlabeled training samples with pseudo labels. Iterations of multi-view clustering and tuning using an updated training set can continue until the updated training set is stable.

Abstract: A foreground attentive neural network is used to learn feature representations. Discriminative features are extracted from the foreground of the input images. The discriminative features are used for various visual recognition tasks such as person re-identification and multi-target tracking. A deep neural network can include a foreground attentive subnetwork, a body part subnetwork and the feature fusion subnetwork. The foreground attentive subnetwork focuses on foreground by passing each input image through an encoder and decoder network. Then, the encoded feature maps are averagely sliced and discriminately learned in the following body part subnetwork. Afterwards, the resulting feature maps are fused in the feature fusion subnetwork. The final feature vectors are then normalized on the unit sphere space and learned by following the symmetric triplet loss layer.

Abstract: A simplified and improved process is described for the production of acrylic fibers, in particular a process for preparing a spinning solution for the production of acrylic fibers.