In the routine production and operation of nuclear power plants (NPPs), safety management is extremely important, and consequently, a fault detection system yielding high accuracy could provide effective decision supp...
详细信息
Antenna arrays are used in many digital signal processing applications due to their ability to locate signal sources. Direction of Arrival (DOA) estimation is a key task of array signal processing. Although various al...
详细信息
Antenna arrays are used in many digital signal processing applications due to their ability to locate signal sources. Direction of Arrival (DOA) estimation is a key task of array signal processing. Although various algorithms have been developed for DOA estimation, their high complexity prevents their use in real-time applications. In this paper, we design and develop an efficient parallel implementation of DOA on DSP which is the most widely used processor in embedded system. Due to the potential parallelism in MUSIC algorithm, it is selected for 2-D DOA estimation. Two computational cores in MUSIC are identified and parallelized. Vectorization of multiple single precision floating point operations is proposed to make good use of the 128-bit vectors on DSP C6678. Then, the parallel DOA estimation algorithm is implemented on one core of DSP C6678 which is the latest version up to now. Experiments are conducted on both 1-D and 2-D antenna array signals. Considerable performance improvement is obtained.
Geological disaster recognition, especially, landslide recognition, is of vital importance in disaster prevention, disaster monitoring and other applications. As more and more optical remote sensing images are availab...
详细信息
Geological disaster recognition, especially, landslide recognition, is of vital importance in disaster prevention, disaster monitoring and other applications. As more and more optical remote sensing images are available in recent years, landslide recognition on optical remote sensing images is in demand. Therefore, in this paper, we propose a deep learning based landslide recognition method for optical remote sensing images. In order to capture more distinct features hidden in landslide images, a particular wavelet transformation is proposed to be used as the preprocessing method. Next, a corrupting & denoising method is proposed to enhance the robustness of the model in recognize landslide features. Then, a deep auto-encoder network with multiple hidden layers is proposed to learn the high-level features and representations of each image. A softmax classifier is used for class prediction. Experiments are conducted on the remote sensing images from Google Earth. The experimental results indicate that the proposed wav DAE method outperforms the state-of-the-art classifiers both in efficiency and accuracy.
Geological disaster recognition on optical image is one of the key techniques in disaster control and disaster relief. Comparing with optical images, remote sensing images contain much higher resolution and more visua...
详细信息
Geological disaster recognition on optical image is one of the key techniques in disaster control and disaster relief. Comparing with optical images, remote sensing images contain much higher resolution and more visualized contents. In this paper, we propose a landslide recognition framework which trains a deep auto-encoder network on the compressed domain. ANN or SVM is used as the classifier for decision making. In addition, in order to meet the requirement of some real-time applications, a high performance training network on CUDA-enabled GPUs is designed and implemented. Experiments are conducted on optical images from Google Earth.
research in the carbon field has experienced rapid growth in recent years and has attracted widespread academic attention. This paper conducts a comprehensive analysis of the carbon field using bibliometric methods. B...
详细信息
Chinese segmentation has attracted amounts of attention in natural language processing in recent years and is the basis of web text mining. The article improved statistics-based method EMI, then we use improved approa...
详细信息
Chinese segmentation has attracted amounts of attention in natural language processing in recent years and is the basis of web text mining. The article improved statistics-based method EMI, then we use improved approach to detect new words in tourism field. The result demonstrates that our method can detect new words significantly, especially in detecting proper nouns and sentiment words which will be helpful in subsequent tasks such as sentiment analysis and word embedding. In additional, this paper analyze parameters which are influential on the effects of new words detection. At last, the article discussed possible application of new word detection in sentiment analysis.
In this paper, we study the problem of learning from label proportions in which label information of data is provided in bag level. In this kind of problem, training data is grouped into various bags and only the prop...
详细信息
In this paper, we study the problem of learning from label proportions in which label information of data is provided in bag level. In this kind of problem, training data is grouped into various bags and only the proportions of positive instances is known. Inspired by proportion-SVM, we propose a new classification model based on twin SVM, which is also in a large-margin framework and only needs to solve two smaller problems. Avoiding making restrictive assumptions of the data, our model can learn the labels of every single instance based on group proportions information. In order to solve the non-convex problem in our new model, we propose an alternative algorithm to obtain the optimal solution efficiently. Also, we prove the effectiveness of our method in theoretical and experimental way.
暂无评论