Beihang University

Microsoft Technology Licensing, LLC

Schneider Electric (china) Co., Ltd.

Juhaokan Technology Co., Ltd.

Feng Lu

Qinping Zhao

Xiaowu Chen

Yunfei Liu

Sijia Wen

Zhimin Wang

Yiwei Bao

Jia Qi Li

Min Liu

Ashley Speicher

Haijun Zhao

Lei Zhang

Haoyun Wu

Jiamin Chen

Hui Wan

Kun Gao

Yang Yu

Zongji Wang

Zhen Ma

Ze Yang

Yuxin Zhao

Anlin Zheng

Thomas Meng

Shuyin Zhao

Jiadong Liang

Mingfei Jiang

Hongye Ren

Yihua Cheng

Kai Li

Jialu Bai

According to an embodiment of the present disclosure, there is provided a solid-state circuit breaker and its self-test method. The solid-state circuit breaker comprises an input end and an output end, the input end being adapted to be connected to a mechanical switch; a first switching device connected between the input end and a first node; a second switching device connected between the first node and the output end; a third switching device and a first resistive element connected in series between the input end and the output end; an auxiliary power branch connected between the first node and a second node; an anti-reverse connection circuit connected between the second node and the input end and between the second node and the output end, and configured to allow a current to flow unidirectionally from the second node to the input end and the output end; a first current sensor configured to sense a current flowing through the second switching device and generate a first current signal; a second current sensor configured to sense a current flowing through the auxiliary power branch and generate a second current signal; and a control unit configured to determine operation states of the first switching device and the second switching device based on the first current signal and the second current signal.

Solid-state circuit breaker and its self-test method

The embodiments of the present disclosure disclose a method for gaze estimation based on physically interpretable features. A specific mode of implementation of this method comprises: constructing a first convolutional neural network, inputting a user's facial image, extracting image features, and regression fitting to obtain an estimated value of the gaze direction; transforming the image features extracted by the first convolutional neural network into physically interpretable features through data processing algorithms, and performing gaze estimation through physically interpretable features; training a gaze estimation model based on physically interpretable features, and on the basis of the trained first convolutional neural network, continuing to optimize the first convolutional neural network using the physically interpretable features.

Method for gaze estimation based on physically interpretable features

Provided are a video call method and a display apparatus. The display apparatus includes: a display, an audio player, a communicator, a user input interface, and a controller. The controller is configured to: control the display to present a first area and a second area on a call interface for a video call, where the first area is for presenting one or more video chatting windows, and the second area is for presenting one or more voice chatting windows; obtain a quantity Y for indicating number of parties currently participating in the video call from the sever; and in response to the quantity Y being greater than a preset quantity X of channels that the display apparatus is able to support, control the display to present X video chatting windows in the first area, and control the display to present (Y-X) voice chatting windows in the second area.

Video call method and display apparatus

The present application provides a gaze point estimation method, device, and an electronic device. The method includes: acquiring user image data; acquiring a facial feature vector according to a preset first convolutional neural network and the facial image; acquiring a position feature vector according to a preset first fully connected network and the position data; acquiring a binocular fusion feature vector according to a preset eye feature fusion network, the left-eye image and the right-eye image; and acquiring position information about a gaze point of a user according to a preset second fully connected network, the facial feature vector, the position feature vector, and the binocular fusion feature vector. In this technical solution, relation between eye images and face images is utilized to achieve accurate gaze point estimation.

Gaze point estimation method, device, and electronic device

The embodiments of the present disclosure disclose gaze estimation cross-scene adaptation method and device based on outlier guidance. A specific implementation of the method comprises: performing pre-training on a source domain based on a given arbitrary gaze estimation model, to obtain a collaborative learning model group; determining an average collaborative model corresponding to each collaborative learning model in the collaborative learning model group, to obtain an average collaborative model group; generating an outlier corresponding to the collaborative learning model group based on a target image, the collaborative learning model group and the average collaborative model group; using an outlier loss function and the outlier to optimize the collaborative learning model group; using any collaborative learning model in the optimized collaborative learning model group to perform gaze estimation. This implementation can effectively stabilize the cross-domain adaptation process and improve the scalability of the system.

Gaze estimation cross-scene adaptation method and device based on outlier guidance

The embodiments of the present disclosure disclose image defogging method. A specific mode of carrying out the method includes the steps to: obtain a polarized fog-free scene image set, generate a first quintuple based on each polarized fog-free scene image, and obtain a first quintuple set; extract pixel coordinates meeting a preset condition from the significant polarization image included in each first quintuple, to obtain a pixel coordinate group to generate a second quintuple; generate a simulated polarization foggy scene image sequence based on each second quintuple, to obtain a simulated polarization foggy scene data set; design a polarization state attention neural network; obtain a polarized foggy scene image, and input the polarized foggy scene image into the polarization state attention neural network to obtain a polarized defogging image. This embodiment can still improve the accuracy of image defogging with poor lighting conditions.

Image defogging method based on simulated polarization foggy scene data set

The embodiments of the present disclosure disclose an edge-guided human eye image analyzing method. A specific implementation of this method comprises: collect a human eye image as an image to be detected; obtain a human eye detection contour map; obtain a semantic segmentation detection map and an initial human eye image detection fitting parameter; performing an iterative search on the initial human eye image detection fitting parameter to determine a target human eye image detection fitting parameter; sending the semantic segmentation detection map and the target human eye image detection fitting parameter as image analyzing results to a display terminal for display. This implementation improves the accuracy at the boundary dividing the pupil-iris area, and increases the structural integrity of the ellipse resulted from dividing the pupil-iris area. In addition, the iterative search can achieve a more accurate ellipse parameter fitting result.

Edge-guided human eye image analyzing method

The embodiments of the present disclosure disclose a method for determining the direction of gaze. A specific implementation of the method includes: obtaining a face or eye image of a target subject, and establishing a feature extraction network; using an adversarial training method to optimize the feature extraction network, and implicitly removing the gaze-irrelevant features extracted by the feature extraction network, so that the feature extraction network extracts gaze-related features from the face or eye image to obtain the gaze-related features; determining the target gaze direction based on the gaze-related features. This implementation can separate the gaze-irrelevant features contained in the image features from the gaze-related features, so that the image features contain the gaze-related features, and that the accuracy and stability of the determined direction of gaze are further improved.

Method for determining the direction of gaze based on adversarial optimization

The embodiments of this disclosure disclose an unsupervised image-to-image translation method. A specific implementation of this method comprises: obtaining an initial image, and zooming the initial image to a specific size; performing spatial feature extraction on the initial image to obtain feature information; inputting the feature information to a style-content separation module to obtain content feature information and style feature information; generating reference style feature information of a reference image in response to obtaining the reference image, and setting the reference style feature information as a Gaussian noise in response to not obtaining the reference image; inputting the content feature information and the reference style feature information into a generator to obtain a target image; and zooming the target image to obtain a final target image. This implementation can be applied to a variety of different high-level visual tasks, and improve the expandability of the whole system.

Unsupervised image-to-image translation method based on style-content separation

The embodiments of the present disclosure disclose method for calculating image highlight reflection separation. A specific implementation of the method includes: controlling a polarized color sensor to shoot a polarized color image, and calculating a polarized chromaticity image; clustering the pixels of diffuse reflection components whose chromaticity difference is within a preset range, to obtain a pixel sequence set; using a robust principal component analysis algorithm for each pixel, to obtain an initial separation result; further processing the polarized color image for highlight reflection separation, to obtain a highlight reflection separation result, and generating a diffuse reflection image. This implementation can achieve highlight reflection separation for images without being affected by the lighting information, and obtain a highlight reflection separation result with higher accuracy through the color and polarization information.

Method for calculating image highlight reflection separation based on polarization guidance

The embodiments of the present disclosure disclose a three-dimensional model's spherical expression calculation method. A specific implementation of the method includes: processing an input three-dimensional model into a dense point cloud model; inputting the dense point cloud model to a multi-level feature extraction module, extracting high-dimensional feature vectors of different abstract levels; inputting the high-dimensional feature vectors, template ball point cloud and abstract preprocessed point cloud model to a point cloud deformation reconstruction module to obtain a deformed reconstruction point cloud model; extracting multi-stage deformation reconstruction process information, combining the multi-stage deformation reconstruction process information with the template ball point cloud to together form complete information describing the three-dimensional model; obtaining a density correspondence from the three-dimensional model to the template ball point cloud and a density correspondence between different three-dimensional models. This implementation does not require time-consuming and labor-intensive manual annotations, improving the efficiency of characterization learning.

Calculation method of three-dimensional model's spherical expression based on multi-stage deformation reconstruction

The embodiments of the present disclosure provide a gaze point calculation method, apparatus and device. The method includes: determining a 3D model of a first scene; obtaining a first panoramic image captured in the first scene, the first panoramic image includes a face image of a face; determining a spatial location of the face according to the first panoramic image; determining a gaze direction corresponding to the face according to the first panoramic image; and determining an eye gaze target according to the 3D model, the spatial location and the gaze direction. The accuracy of tracking the eye gaze in 3D model is improved.

Gaze point calculation method, apparatus and device

The present invention proposes method of gaze estimation with 3D face reconstructing, which precisely reconstructs a 3D face to accurately calculate the gaze direction of the face. The present invention decomposes the gaze direction calculation task into two sub-tasks of gaze-aligned 3D face reconstruction, and gaze contact realized by rotation. First, a 3D morphable model is used to reconstruct a basic 3D face with a 2D reference face image as input. Then, shape adjustment is performed on the basic 3D face to obtain a shape-aligned 3D reconstructed face. Finally, a gaze-aligned 3D face is obtained by replacing the textures of the shape-aligned 3D face eye area. For the subtask of gaze contact, the present invention proposes a novel method for realizing gaze contact by rotating a 3D face, which does not depend on too many assumptions and can be widely used in the field of virtual reality.

Method of gaze estimation with 3D face reconstructing

The embodiments of the present disclosure disclose an augmented reality system that supports customized multi-channel interaction. One embodiment of the augmented reality system comprises: a head-mounted sensor assembly, a computing device, and a display module; the head-mounted sensor assembly is used to capture the user's multi-channel interactive input information and transmit the interactive input information to the computing device; the computing device is used to generate or modify the display content of the augmented reality according to the interactive input information; the display module is used to overlay display the background content with the display content of the augmented reality. The augmented reality system, by arranging the display module to the far end of the head-mounted sensor assembly, can simplify the structure of the head-mounted sensor assembly, and reduce the weight of the head-mounted sensor assembly, providing convenience for installing other sensors.

Augmented reality system supporting customized multi-channel interaction

This disclosure provides a method and an apparatus for monitoring a working state, which automatically collect an image of a staff in real time, determines point of gaze information of the staff based on the image of the staff thus collected, and further determines the working state of the staff according to the point of gaze information. Since the whole process does not require the participation of the staff, the normal work of the staff is not disturbed. Moreover, the accuracy in the monitoring of the working state is improved by avoiding influence of the subjective factors on the assessment result if staff participation is involved.

Growing community of inventors

Beijing, China

Feng Lu