Abstract: Disclosed are a facial beauty prediction method and device based on multi-task migration. The method includes: performing similarity measurement based on a graph structure on a plurality of tasks to obtain an optimal combination of the plurality of tasks; constructing a facial beauty prediction model including a feature sharing layer based on the optimal combination; migrating feature parameters of an existing large-scale facial image network to the feature sharing layer of the facial beauty prediction model; inputting facial images for training to pre-train the facial beauty prediction model; and inputting a facial image to be tested to the trained facial beauty prediction model to obtain facial recognition results.

Abstract: A method for measuring an antenna downtilt angle based on a multi-scale deep semantic segmentation network is disclosed, including: collecting base station antenna data by using an unmanned aerial vehicle, and labeling an acquired antenna image with a labeling tool to make a data set; calling the data set for training and debugging a model; recognizing and detecting a target antenna, performing semantic segmentation on an output image, finally obtaining a target image finally segmented, and calculating a downtilt angle of the target image. The method is highly applicable, cost-effective, and safe.

Abstract: Disclosed are a method and apparatus for SAR image data enhancement, and a storage medium. The method includes: processing an SAR target image by electromagnetic simulation to acquire an SAR electromagnetic simulation image; and processing the SAR electromagnetic simulation image and the SAR target image by a generative adversarial network to obtain a set of virtual samples of the SAR target image.

Abstract: Disclosed is a method for vein recognition, the method includes: performing a difference operation and a channel connection on two to-be-verified target vein images respectively to obtain a difference image and a two-channel image of the two target vein images; performing the channel connection on the obtained difference image and two-channel image to obtain a three-channel image, so as to use the three-channel image as an input of a CNN network; fine-tuning a pre-trained model SqueezeNet that completes training on an ImageNet; integrating the difference image and the three-channel image through a cascade optimization framework to obtain a recognition result; regarding a pair of to-be-verified images as a sample, transforming the sample, taking the transformed sample as the input of the CNN network, obtaining a recognition result by supervised training on the network.

Abstract: The present disclosure discloses an error modeling method and device for prediction context of reversible image watermarking. A predictor based on omnidirectional context is established; then, the prediction context is self-adaptively error modeled to obtain a self-adaptive error model; and finally, output data from the self-adaptive error model is fed back to the predictor to update and correct the prediction context, so as to correct a prediction value of a current pixel x[i,j]. Since the non-linear correlation between the current pixel and the prediction context thereof, i.e., the non-linear correlation redundancy between pixels can be found by the error modeling of the prediction context of the predictor, the non-linear correlation redundancy between the pixels can be effectively removed. Thus, the embeddable watermarking capacity can be increased.

Abstract: Disclosed are a facial beauty prediction method and device based on multi-task migration. The method includes: performing similarity measurement based on a graph structure on a plurality of tasks to obtain an optimal combination of the plurality of tasks; constructing a facial beauty prediction model including a feature sharing layer based on the optimal combination; migrating feature parameters of an existing large-scale facial image network to the feature sharing layer of the facial beauty prediction model; inputting facial images for training to pre-train the facial beauty prediction model; and inputting a facial image to be tested to the trained facial beauty prediction model to obtain facial recognition results.

Abstract: A double attention network (DANet)-based drone patrol and inspection system for coastline floating garbage, including: an image acquisition module configured to shoot a video of a coastline in need of patrol and inspection by using a drone, and obtain an image from the video; a feature extraction module configured to extract shallow features and deep features, fuse the shallow features and the deep features to obtain a shared feature, and finally output a panoramic recognition result; a network training module configured to perform training on the labeled image so that the network can recognize the coastline and floating garbage; and a path correction module configured to adjust a flying direction of the drone.

Abstract: A method and a device for acquiring three-dimensional coordinates of ore based on mining process are disclosed. The method includes: obtaining a two-dimensional coordinate of the ore by using a YOLACT algorithm and a NMS algorithm to obtain a prediction mask map, obtaining depth information of the ore based on the color map and the infrared depth map, and combining the two-dimensional coordinate with the depth information to obtain a three-dimensional coordinate of the ore.

Abstract: A method and system for mobile communication base station antenna measurement is disclosed. The method comprises steps of: acquiring a set of images containing antennas of a base station; processing the set of images with a model based on instance segmentation network, and generating visualized images corresponding to the set of images of antennas; calculating, from the visualized images, the quantity of antennas of the base station and separating data for each antenna; measuring parameters of each antenna by data fitting. The system comprises a processor and a memory storing program instructions thereon, the program instructions executable by the processor to cause the system to perform the steps of the method.

Abstract: A method for predicting a face beauty grade includes the following steps of: acquiring a beautiful face image of a face beauty database, preprocessing the beautiful face image, and extracting a beauty feature vector of the beautiful face image, the preprocessing unifying data of the beautiful face image; recognizing continuous features of samples of the same type in a feature space by using a bionic pattern recognition model, and classifying the beauty feature vector to obtain a face beauty grade prediction model; and collecting a face image to be recognized, and inputting the face image to be recognized into the face beauty grade prediction model to predict a face beauty grade and obtain the beauty grade of the face image to be recognized.

Abstract: A double attention network (DANet)-based drone patrol and inspection system for coastline floating garbage, including: an image acquisition module configured to shoot a video of a coastline in need of patrol and inspection by using a drone, and obtain an image from the video; a feature extraction module configured to extract shallow features and deep features, fuse the shallow features and the deep features to obtain a shared feature, and finally output a panoramic recognition result; a network training module configured to perform training on the labeled image so that the network can recognize the coastline and floating garbage; and a path correction module configured to adjust a flying direction of the drone.

Abstract: A method for measuring an antenna downtilt angle based on a multi-scale deep semantic segmentation network is disclosed, including: collecting base station antenna data by using an unmanned aerial vehicle, and labeling an acquired antenna image with a labeling tool to make a data set; calling the data set for training and debugging a model; recognizing and detecting a target antenna, performing semantic segmentation on an output image, finally obtaining a target image finally segmented, and calculating a downtilt angle of the target image. The method is highly applicable, cost-effective, and safe.

Abstract: Disclosed are a method and an apparatus for multi-scale SAR image recognition based on attention mechanism. According to the method, a whole image recognition network is adjusted by training a SAR training image by an attention prediction subnet, a region-of-interest positioning subnet and an image classification subnet in combination with a network loss, which greatly improves a network performance; and in addition, an attention prediction map is generated by attention mechanism to position a most prominent feature part in the SAR image, which greatly eliminates a redundancy of image features in a machine vision, effectively determines a region-of-interest, reduces interference of image noises, greatly reduces an image processing time, improves a target recognition accuracy, is beneficial to next target positioning, and has a significant improvement on a network recognition speed integrally.

Abstract: Disclosed are method and apparatus for SAR image recognition based on multi-scale features and broad learning. A region of interest of an original SAR image is extracted by centroid localization, the image is rotated and added with noise for enhancing the data volume, the image is downsampled, LBP features and PPQ features are extracted, an LBP feature vector XLBP and an LPQ feature vector XLPQ are cascaded to achieve dimension reduction by principal component analysis to obtain a fusion feature data Xm, the fusion feature data Xm is input to a broad learning network for image recognition and a recognition result is output. By fusing the LBP features and the LPQ features, complementary information is fully utilized and redundant information is reduced. The broad learning network is used to improve the training speed and reduce the time cost. As a result, the recognition effect is more stable, robust and reliable.

Abstract: Disclosed is a method for vein recognition, the method includes: performing difference operation and channel connection on two to-be-verified target vein images respectively to obtain a difference image and a two-channel image of the two target vein images; performing channel connection on the obtained difference image and two-channel image to obtain a three-channel image, so as to use the three-channel image as an input of a CNN network; fine-tuning a pre-trained model SqueezeNet that completes training on an ImageNet; integrating the difference image and the three-channel image through a cascade optimization framework to obtain a recognition result; regarding a pair of to-be-verified images as a sample, transforming the sample, taking the transformed sample as the input of the CNN network, obtaining a recognition result by supervised training on the network, which solve the problem of less training data and improve the accuracy of recognition with a small network storage capacity.

Abstract: An intelligent medical material supply robot based on Internet of Things and SLAM technology is disclosed, which realizes localization and mapping through a binocular camera and a lidar. A cloud data center schedules the medical material supply robot in real time according to material usage. The material supply robot receives corresponding scheduling information, and according to localization of the robot and map information, dynamically avoids obstacles by using a path planning algorithm to go to a designated floor for materials delivery.

Abstract: A method for automatic shopping in a shopping mall, a storage medium, an electronic device, and a device are disclosed. After a consumer selects commodities from an input means, a path along which a moving means needs to travel to pick up the commodities from a shelf is planned on a global map according to first coordinates carried in commodity information, then the moving means is controlled to automatically walk in the shopping mall according to the planned path, the height of lifting means is adjusted according to height values of the corresponding commodities on the shelf, and when the moving means arrives at designated positions on the path, a pickup means on the lifting means picks up the commodities selected by the consumer from the shelf.

Abstract: A method and a device for acquiring three-dimensional coordinates of ore based on mining process are disclosed. The method includes: obtaining a two-dimensional coordinate of the ore by using a YOLACT algorithm and a NMS algorithm to obtain a prediction mask map, obtaining depth information of the ore based on the color map and the infrared depth map, and combining the two-dimensional coordinate with the depth information to obtain a three-dimensional coordinate of the ore.

Abstract: Disclosed are an education assisting robot and a control method thereof. The method includes: capturing and recognizing students' faces from shot images, and checking students' attendance; and capturing a teacher's face from the shot images and identifying a target, and target-following the teacher. The role of a target character can be automatically distinguished from the images, and different actions can be made for different target characters, including attendance checking and target following, so as to provide more different response functions and reduce the workload of teachers.

Abstract: A method and system for mobile communication base station antenna measurement is disclosed. The method comprises steps of: acquiring a set of images containing antennas of a base station; processing the set of images with a model based on instance segmentation network, and generating visualized images corresponding to the set of images of antennas; calculating, from the visualized images, the quantity of antennas of the base station and separating data for each antenna; measuring parameters of each antenna by data fitting. The system comprises a processor and a memory storing program instructions thereon, the program instructions executable by the processor to cause the system to perform the steps of the method.