In order to solve the key technical problem in the verification process of mechanical and electrical coal mine anemometer, the paper presents digital recognition approach base on the characters of the anemometer image...
详细信息
As the complexity of space exploration missions augments,how to enhance the overall performance of communication,ranging or other functions has become a challengeable *** the integration of communication and ranging,w...
详细信息
As the complexity of space exploration missions augments,how to enhance the overall performance of communication,ranging or other functions has become a challengeable *** the integration of communication and ranging,we present a bit-level composite signal for simultaneous ranging and *** this composite method,through a specially designed mapping scheme using low-weight codewords,the information sequence is converted to a sparse sequence which is then superimposed on the ranging *** ranging,the correlation characteristics of the ranging code component can be maintained to calculate the transmitter-receiver *** communications,the sparse sequence can be extracted without interference by eliminating the ranging code *** results show that the proposed composite signal can support communication and ranging simultaneously with limited sacrifice of ranging performance,and the performance loss of ranging can be controlled and minimized by lowering the density of information sequences using different sparsification encoding methods.
Short videos on the Internet have a huge amount, but most of them are unlabeled. In this paper, a rough short video labelling method based on the image classification neural network is proposed. Convolutional auto-enc...
详细信息
Short videos on the Internet have a huge amount, but most of them are unlabeled. In this paper, a rough short video labelling method based on the image classification neural network is proposed. Convolutional auto-encoder is applied to train and learn unlabeled video frames, in order to obtain feature in the specific level. With these features, the video key-frames are extracted by the feature clustering method. These key-frames which represent the video content are put into an image classification network, so that the labels of every video clip can be got. In addition, the different architectures of convolutional auto-encoder are estimated, and a better performance architecture through the experiment result is selected. In the final experiment, the video frame features from the convolutional auto-encoder are compared with those from other extraction methods, where it illustrates remarkable results by the proposed method.
The paper inctroduces the guide for the design of a portable instrument for straightness measurement of large machine tool. the measuring instrument is designed with two-quadrant photocell which is used as the positio...
详细信息
A rate-compatible spatially coupled repeat-accumulate (RC-SC-RA) code is proposed. Its protograph is obtained by extending a given (J, K, L) SC-RA coupled chain (denoted as the mother chain) with extra check nodes and...
详细信息
A rate-compatible spatially coupled repeat-accumulate (RC-SC-RA) code is proposed. Its protograph is obtained by extending a given (J, K, L) SC-RA coupled chain (denoted as the mother chain) with extra check nodes and parity bit nodes T times. At each time, the extension is realized via coupling the message bits in the same way as that in the mother chain. Rate-compatibility is achieved by adjusting the extension parameters and applying random puncturing technique. Density evolution analysis shows that the iterative decoding thresholds of all the member codes in the proposed RC-SC-RA code family are very close to Shannon limits over the binary erasure channel. Finite length simulation results are consistent with the thresholds well. Moreover, the proposed RC-SC-RA codes perform better than spatially coupled low density parity check (SC-LDPC) codes in decoding performance especially in lower-rate region.
The granary monitoring systems in used in China are mostly based on wired networks such as RS-485 serial bus, CAN fieldbus, etc. Inevitably these monitoring systems have drawbacks of high installation cost, high failu...
详细信息
An important issue in the classification of electromyography signals is the classifier design. For surface electromyography signals in the acquisition process of serious interference by the noise, in particular freque...
详细信息
ABSTRACTCracks are an important indicator of pavement health, and it is difficult to achieve pixel-level segmentation of small and thin cracks. The existing network often experiences false segmentation and missed segm...
详细信息
ABSTRACTCracks are an important indicator of pavement health, and it is difficult to achieve pixel-level segmentation of small and thin cracks. The existing network often experiences false segmentation and missed segmentation. Accordingly, a novel end-to-end U-Net-like full convolutional crack segmentation network is constructed. First, we propose a multi-layer feature fusion module to aggregate the texture and semantic features at each stage of encoder, so that the network can find smaller and thinner crack. Second, we design a novel residual structure with a pointwise convolution. Each stage of the encoder and decoder incorporates a residual structure to facilitate the fusion of feature maps with different spatial dimensions. It can also prevent the gradient vanish in the network training process. Finally, we utilise the maximum unpooling to restore spatial structure in up-sampling, which exploits the indices of maximum feature value in down-sampling. Therefore, high-frequency information is better preserved to help accurately restore the details of crack edges. To verify the proposed network performance, experiments are carried out on four open datasets, the proposed network can achieve better performance among five classical networks.
Most target tracking is based on a lot of samples training to build the model of the target, which is then carried on the tracking processing. This will need to choose a lot of tracked target samples for learning and ...
详细信息
Learning-based multi-task models have been widely used in various scene understanding tasks,and complement each other,i.e.,they allow us to consider prior semantic information to better infer *** boost the unsupervise...
详细信息
Learning-based multi-task models have been widely used in various scene understanding tasks,and complement each other,i.e.,they allow us to consider prior semantic information to better infer *** boost the unsupervised monocular depth estimation using semantic segmentation as an auxiliary *** address the lack of cross-domain datasets and catastrophic forgetting problems encountered in multi-task training,we utilize existing methodology to obtain redundant segmentation maps to build our cross-domain dataset,which not only provides a new way to conduct multi-task training,but also helps us to evaluate results compared with those of other *** addition,in order to comprehensively use the extracted features of the two tasks in the early perception stage,we use a strategy of sharing weights in the network to fuse cross-domain features,and introduce a novel multi-task loss function to further smooth the depth *** experiments on KITTI and Cityscapes datasets show that our method has achieved state-of-the-art performance in the depth estimation task,as well improved semantic segmentation.
暂无评论