Semi-supervised learning is adopted for fine-grained image classification with convolutional networks. Compared with the traditional approach of distillation, we obtain accuracy improvement of ~3 percent points under ...
详细信息
ISBN:
(纸本)9781943580705
Semi-supervised learning is adopted for fine-grained image classification with convolutional networks. Compared with the traditional approach of distillation, we obtain accuracy improvement of ~3 percent points under the upper limit of supervised learning on cassava-disease dataset.
Deep learning methods can enhance the efficiency of tumor segmentation in breast ultrasound (BUS) images. However, noise interference, small tumors, and blurred boundaries can reduce segmentation accuracy. We design a...
Deep learning methods can enhance the efficiency of tumor segmentation in breast ultrasound (BUS) images. However, noise interference, small tumors, and blurred boundaries can reduce segmentation accuracy. We design a three-branch challenge-aware U-net (CAU-net) to address these main challenges in BUS images. Our CAU-net extracts the features from three challenge-aware encoders in parallel first. Secondly, we propose an adaptive aggregation layer (AAL) to merge the multi-scale features of three challenging branches, enabling the network to adaptively handle different breast lesion samples with these main challenges. To further enhance the accuracy of segmentation, we introduce the graph reasoning module (GRM) to the network to model the correlation between the channels of the features and acquire the global information in the features. The result of our experiment on two datasets demonstrates the superiority of CAU-net over the advanced medical image segmentation methods. Our code can be downloaded from https://***/tzz-ahu .
Since the marine diesel engine is one of the most popular power equipments for modern shipping, accurate and timely diagnose the faults occurred in diesel engine is extraordinary important for long service life and hi...
详细信息
In order to periodically reassess the status of the alternate path route(APR)set and to improve the efficiency of alternate path construction existing in most current alter-nate path routing protocols,we present a cro...
详细信息
In order to periodically reassess the status of the alternate path route(APR)set and to improve the efficiency of alternate path construction existing in most current alter-nate path routing protocols,we present a cross-layer design and ant-colony optimization based load-balancing routing protocol for ad-hoc networks(CALRA)in this *** CALRA,the APR set maintained in nodes is aged and reas-sessed by the inherent mechanism of pheromone evaporation of ant-colony optimization algorithm,and load balance of network is achieved by ant-colony optimization combining with cross-layer synthetic *** efficiency of APR set construction is improved by bidirectional and hop-by-hop routing update during routing discovery and routing maintenance ***,ants in CALRA deposit simulated pheromones as a function of multiple parameters corresponding to the information collected by each layer of each node visited,such as the distance from their source node,the congestion degree of the visited nodes,the current pheromones the nodes possess,the velocity of the nodes,and so on,and provide the information to the visiting nodes to update their pheromone tables by endowing the different parameters corresponding to different information and different weight values,which provides a new method to improve the congestion problem,the shortcut problem,the convergence rate and the heavy overheads commonly existed in existing ant-based routing protocols for ad-hoc *** performance of the algorithm is measured by the packet delivery rate,good-put ratio(routing overhead),and end-to-end *** results show that CALRA performs well in decreasing the route overheads,balancing traffic load,as well as increasing the packet delivery rate,etc.
The current automatic decoding method of the Morse telegram has limited accuracy, and can't adapt to signal distortion and code length deviation of the manual telegram. This paper introduces the deep learning meth...
详细信息
ISBN:
(数字)9781728161068
ISBN:
(纸本)9781728161075
The current automatic decoding method of the Morse telegram has limited accuracy, and can't adapt to signal distortion and code length deviation of the manual telegram. This paper introduces the deep learning method and constructs an automatic decoding model, which integrates feature extraction, sequence modeling and transcription into an end-to-end training neural network. The time-frequency diagrams of signals are used for training and testing. Experimental results show that the decoding system has strong adaptability to manual deviation and frequency drift, and is robust in a noisy environment.
The aim of our work is to investigate the use of two powerful feature descriptor known as Zernike moments and histogram of oriented gradient (HOG) for facial images extracted from a video sequence uttered ten times by...
详细信息
In this paper, we analyzed the blindness of traditional clustering algorithms, which select cluster head based on residual energy. Then we proposed the Dynamic Clustering Algorithm (DCA) in Mobile Wireless Sensor Netw...
详细信息
Human salience of pedestrians images is distinctive and has been shown importantly in person re-identification (or pedestrians identification) problem. Thus, how to obtain the salient area of pedestrian images is impo...
详细信息
Technological innovation is becoming one of the critical factors in promoting social development all over the world. The vigorous development of patent applications in recent years provides an opportunity to reveal th...
详细信息
In an anonymous secret sharing scheme, the secret can be reconstructed without knowledge of which participant hold which share. That is, in such scheme the secret can be recovered from the shares without the identitie...
详细信息
暂无评论