The estimation of image geometry benefits many applications in the field of computer vision, such as stereo correspondence, 3-D reconstruction, and camera self-calibration. It is very challenging since the proportion of inliers in putative correspondences is usually very low, and traditional image geometry estimation methods (such as Ransac) suffer from low accuracy at a high outlier ratio. In this paper, we tackle the two-view image geometry estimation problem and propose a new robust estimation method L 2 E-LSC (short for L 2 E with local structure constraint) based on the L 2 E algorithm. In particular, we first establish initial correspondences by feature description matches, and then estimate the fundamental matrix and homography using L 2 E-LSC and get the refined correspondences. The L 2 E-LSC is able to robustly deal with the noise and outliers contained in point correspondences. Extensive experiments conducted on real images from public available datasets have demonstrated that it can achieve good estimation accuracy and superior performance over previous approaches, particularly when there are severe outliers.
This paper presents a real-time vision based robot teleoperation system that consists of a three-dimensional (3D) vision subsystem and a slave robot which are connected by LAN. The vision subsystem utilizes an Asus Xtion Pro Live camera to get the 3D data of the operation scene. The vision system is used to determine the position and orientation of a four-ball feature frame held by the operator. Then the position and the orientation are used to control a remote robot. In the vision subsystem, support vector domain description (SVDD) is adopted to detect the balls on the feature frame. In this paper, we propose a novel colour table to speed up the detection procedure and utilize Kalman filter for ball tracking to reduce the detection area for further acceleration. The operator can see the motion of the robot which enables the operator make some corrections constantly. The SVDD colour classifier and the teleoperation system are tested in experiments.
Recently, convolutional neural networks (CNNs) have been widely employed to promote the face hallucination due to the ability to predict high-frequency details from a large number of samples. However, most of them fail to take into account the overall facial profile and fine texture details simultaneously, resulting in reduced naturalness and fidelity of the reconstructed face, and further impairing the performance of downstream tasks (e.g., face detection, facial recognition). To tackle this issue, we propose a novel external-internal split attention group (ESAG), which encompasses two paths responsible for facial structure information and facial texture details, respectively. By fusing the features from these two paths, the consistency of facial structure and the fidelity of facial details are strengthened at the same time. Then, we propose a split-attention in split-attention network (SISN) to reconstruct photorealistic high-resolution facial images by cascading several ESAGs. Experimental results on face hallucination and face recognition unveil that the proposed method not only significantly improves the clarity of hallucinated faces, but also encourages the subsequent face recognition performance substantially. Codes have been released at this https URL.
To improve the performance of high-frequency information acquisition data processing and meet the differentiated power service requirements in low-voltage distribution grid, this paper first investigates an edge processing architecture for high-frequency information acquisition in low-voltage distribution grid. Then, an edge processing mechanism for high-frequency information acquisition is proposed to optimize the high-frequency information acquisition data distribution strategy and communication and computation resource allocation strategies. Simulation results demonstrate that the proposed mechanism achieves better data processing performances compared with traditional data processing mechanism
The existing occlusion face recognition algorithms almost tend to pay more attention to the visible facial components. However, these models are limited because they heavily rely on existing face segmentation approaches to locate occlusions, which is extremely sensitive to the performance of mask learning. To tackle this issue, we propose a joint segmentation and identification feature learning framework for end-to-end occlusion face recognition. More particularly, unlike employing an external face segmentation model to locate the occlusion, we design an occlusion prediction module supervised by known mask labels to be aware of the mask. It shares underlying convolutional feature maps with the identification network and can be collaboratively optimized with each other. Furthermore, we propose a novel channel refinement network to cast the predicted single-channel occlusion mask into a multi-channel mask matrix with each channel owing a distinct mask map. Occlusion-free feature maps are then generated by projecting multi-channel mask probability maps onto original feature maps. Thus, it can suppress the representation of occlusion elements in both the spatial and channel dimensions under the guidance of the mask matrix. Moreover, in order to avoid misleading aggressively predicted mask maps and meanwhile actively exploit usable occlusion-robust features, we aggregate the original and occlusion-free feature maps to distill the final candidate embeddings by our proposed feature purification module. Lastly, to alleviate the scarcity of real-world occlusion face recognition datasets, we build large-scale synthetic occlusion face datasets, totaling up to 980193 face images of 10574 subjects for the training dataset and 36721 face images of 6817 subjects for the testing dataset, respectively. Extensive experimental results on the synthetic and real-world occlusion face datasets show that our approach significantly outperforms the state-of-the-art in both 1:1 face verification and 1:N face identification.
Prototype learning is extensively used in few-shot semantic segmentation due to its excellent capability of semantic information extraction and effective prevention of overfitting. The previous prototype based methods ignore the frequency discrepancy inside the object, thereby leading to semantic confusion of the object. In this paper, we propose a frequency guided network (FGNet) which explicitly models the semantic information of different frequencies and precisely guides the semantic alignment of the object. Specifically, the proposed FGNet consists of two modules: a frequency separation module (FSM) and a multi-guided feature enrichment module (MG-FEM) to complete the multi-frequency semantic information extraction and alignment, respectively. Experiments on PASCAL- $5^{i}$ dataset show that our FGNet achieves mIoU score of 61.2% in 1-shot which surpasses the state-of-the-art methods.
Lower-limb rehabilitation robots can support the training of patients with neurological gait disorders. Compared to the robot-driven control strategy which permits patients to remain passive during the training, Patient-driven control strategy can motivate the patient's participation and accelerate the rehabilitation process. In this paper, based on a selfdeveloped lower-limb rehabilitation robot, a patient-driven control method is developed and implemented to realize a patient-in-charge gait training strategy. During the training, the proposed control algorithm adapts the motion pattern of the robot according to the estimated patient’s voluntary efforts and increases the patient’s movement freedom by a certain amount of robot compliance. By adjusting the parameters, the degree of robot compliance can be set based on the patient’s voluntary abilities. Experiment results are presented to evaluate the basic principal and technical function of the proposed method and show positive for improving the subject’s voluntary contribution during the training.
Abstract Impulse noise is regarded as an outlier in the local window of an image. To detect noise, many proposed methods are based on aggregated distance, including spatially weighted aggregated distance, n nearest neighbour distance, local density, and angle‐weighted quaternion aggregated distance. However, these methods ignore the weight of each pixel or have limited adaptability. This study introduces the concept of degree of aggregation and proposes a weighting method to obtain the weight vector of the pixels by minimizing the degree of aggregation. The weight vector obtained gives larger components on the signal pixels than on the noisy pixels. Then it is fused with the aggregated distance to form a weighted aggregated distance that can reasonably characterise the noise and signal. The weighted aggregated distance, along with an adaptive segmentation method, can effectively detect the noise. To further enhance the effect of noise detection and removal, an adaptive selection strategy is incorporated to reduce the noise density in the local window. At last, noisy pixels detected are replaced with the weighted channel combination optimization values. The experimental results exhibit the validity of the proposed method by showing better performance in terms of both objective criteria and visual effects.
In this paper, we propose a manifold learning based algorithm using 'Nearest Feature Line - NFL' to hallucinate high-resolution face image. According to the fact that existing NFL can effectively characterize the geometrical proportions to the face samples, we propose using NFL metric to define the neighborhood relations between face samples. Our algorithm can solve the problem that traditional method cannot effectively reveal the similar local geometry between high-resolution and low-resolution face manifolds under the condition that the training sample size is small. Moreover, in order to enhance the representation capacity of available face samples and reduce the computational complexity, we select neighborhood samples for each input LR image. Experimental results demonstrate that our algorithm can generates clearer local feature details, and the PSNR is 1.4 dB higher than that of the best manifold learning based method reported so far.
Geographical Information System - GIS has strong power in handle, orientation and communication in geographical information, the development and maturity of GIS technology allow that logistics distribution can use it to analyze spatial network and track the distribution process. First, the basic model of logistics distribution system is introduced, then the framework of logistics distribution system based on GIS, the module and function design about the system are discussed in the paper. This framework in the practical application has made impressive result.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)