|
|
|
题名
|
作者
|
年代
|
出处
|
被引量
|
| 1 | Shallow Convolutional Neural Networks for Acoustic Scene Classification显示文摘最近,深神经的网络,它包括 convolutional 神经网络(CNN ) ,广泛地被用于声学的景色分类(ASC ) 。由一些简化 CNN 在深 CNN 上看了改进的事实激发了,例如视觉几何学组网(VGG 网) ,我们弄明白怎么与改进性能简化 VGG 网风格体系结构到浅 CNN。分享的马克斯和正规化也是的批申请了更好的精确性。与声学的场面和事件(DCASE ) 的察觉和分类上的一系列控制测试, 2016 数据设定,我们的浅 CNN 完成 6.7% 改进,并且把时间复杂性归结为 5% ,与 VGG 网风格 CNN 相比。 | LU Lu YANG Yuhong JIANG Yuzhi AI Haojun TU Weiping | 2018 | Wuhan University Journal of Natural Sciences2018，23，2： | 3 |
| 2 | Introduction to AVS audio显示文摘 | Ai Haojun Chen Shuixian Hu Ruimin | 2006 | J Conmpute Sci and Technol2006，21，3： | 1 |
| 3 | Selective Image Matting with Scalable Variance and Model Rectification显示文摘Bayesian Matting has four limitations.Firstly,Bayesian matting makes strong assumption that the texture distribution of nature image satisfies Gaussian distribution with fixed variance.This assumption will fail for complex texture distribution.In order to extract the nature images with complex texture distribution,we design an information entropy approach to estimate the scalable variance.Secondly,when the opacity is near the boundary of the value range,Bayesian matting method may be failure because of the error computation of opacity.Therefore,a rectification approach is proposed to adjust the computation model and keep the opacity within the valid value range.Thirdly,Bayesian matting is a local sample method which may miss some valid samples of matting.We propose a selection function to integrate valid global sample matting result into above matting framework as a supplement to the local sample matting result.The proposed function is compose of three criteria,that is,the similarity of results,the overlapping degree of samples,and the similarity of neighborhood.Fourthly,in order to obtain a smooth and vivid matte,the result is further refined by considering correlation between neighbouring pixels.Finally,We use online benchmark for image matting to evaluate the proposed method with both qualitative observation and quantitative analysis.The experiments show that our method achieves a competitive advantages over other methods. | Xiao Chen Fazhi He Yiteng Pan Haojun Ai | 2017 | 国际计算机前沿大会会议论文集2017，，1： | 0 |
| 4 | Development and implementation of a dynamic and 4D GIS based on semantic location model显示文摘In the big data era,robust solutions are obliged to be proposed to integrate and represent data from different formats and with different contents to assist the decision-making.Current cartographic and geographic information systems have limited capabilities for solving these problems.This paper describes an automatic and comprehensive system that conducts data fusion from all potentially related sources.In this system,a new Semantic Location Model(SemLM)is established to present the semantic concepts and location feature and demonstrate how locations are interrelated.In the SemLM,various types of location descriptors in different application scenarios can be analyzed and understood.Additionally,considering the challenges involved in data-intensive computation and visualization,this paper implements a Place-based Pan-Information System(P2S)as an innovative 4D system that dynamically associates and visualizes place-based information,using public security as the case study. | Xinyan Zhu Tao Hu Xinyue Ye Wei Guo Liang Huang Hanjiang Xiong Haojun Ai Bing She Qing Xiong Lian Duan | 2019 | Geo-Spatial Information Science2019，22，3： | 0 |
您所在的机构暂未开通此部分的功能，请联系管理员