The intensive computations in convolutional neural networks (CNNs) pose challenges for resource-constrained devices; eliminating redundant computations from convolution is essential. This paper gives a principled meth...
ISBN:
(纸本)9781713871088
The intensive computations in convolutional neural networks (CNNs) pose challenges for resource-constrained devices; eliminating redundant computations from convolution is essential. This paper gives a principled method to detect and avoid transient redundancy, a type of redundancy existing in input data or activation maps and hence changing across inferences. By introducing a new form of convolution (TREC), this new method makes transient redundancy detection and avoidance an inherent part of the CNN architecture, and the determination of the best configurations for redundancy elimination part of CNN backward propagation. We provide a rigorous proof of the robustness and convergence of TREC-equipped CNNs. TREC removes over 96% computations and achieves 3.51× average speedups on microcontrollers with minimal (about 0.7%) accuracy loss.
At the early stage of software lifecycle,the complexity measurement of UML class diagrams plays an important role in software development,testing and maintenance,and provides guidance for developing high quality *** o...
详细信息
At the early stage of software lifecycle,the complexity measurement of UML class diagrams plays an important role in software development,testing and maintenance,and provides guidance for developing high quality *** order to study which one is better,simple or complex metrics,this paper analyzes and compares four typical metrics of UML class diagrams from experimental software engineering view ***,analyzability and maintainability were classified and predicted for 27 class diagrams related to a banking system by means of algorithm C5.0 within the famous toolkit *** suggest that the performance of simple metrics is not inferior to that of complex metrics,in some cases even better than that of complex metrics.
What-if analysis can provide more meaningful information than classical OLAP. Multi-scenario hypothesis based on historical data needs efficient what-if data view support. In general, delta table for what-if analysis ...
详细信息
This volume contains papers selected for presentation at the 7th Asia Pacific Conference on Web Technology (APWeb 2005), which was held in Shanghai, China during March 29–April 1, 2005. APWeb is an international conf...
详细信息
ISBN:
(数字)9783540318491
ISBN:
(纸本)9783540252078
This volume contains papers selected for presentation at the 7th Asia Pacific Conference on Web Technology (APWeb 2005), which was held in Shanghai, China during March 29–April 1, 2005. APWeb is an international conference series on WWW technologies and is the primary forum for researchers and practitioners from both academia and industry to exchange knowledge on WWW-related technologies and new advanced applications. APWeb 2005 received 420 submissions from 21 countries and regions worldwide, including China, Korea, Australia, Japan, Taiwan, France, UK, Canada, USA, India, Hong Kong, Brazil, Germany, Thailand, Singapore, Turkey, Spain, Greece, Belgium, New Zealand, and UAE. After a thorough review process for each submission by the Program Committee members and expert reviewers recommended by PC members, APWeb 2005 accepted 71 regular research papers (acceptance ratio 16.9%) and 22 short papers (acceptance ratio 5.2%). This volume also includes 6 keynote papers and 11 invited demo papers. The keynote lectures were given by six leading experts: Prof. Ah Chung Tsoi (Australia Research Council), Prof. Zhiyong Liu (National Nature Science Foundation of China), Prof. John Mylopoulos (University of Toronto), Prof. Ramamohanarao (Rao) Kotagiri (University of Melbourne), Prof. Calton Pu (Georgia Tech), and Prof. Zhiwei Xu (Chinese Academy of Sciences).
Most of the existing slam algorithms are designed based on the assumption of a static environment, this strong assumption limits the practical application of most slam systems. The main reason is that moving objects w...
Most of the existing slam algorithms are designed based on the assumption of a static environment, this strong assumption limits the practical application of most slam systems. The main reason is that moving objects will cause feature mismatch in the pose estimation process, which in turn affects the accuracy of localization and mapping. In this paper, we propose a SLAM algorithm in a dynamic environment. First, we use the BlendMask network to detect potential moving objects to generate masks for dynamic objects. The geometrically constrained joint optical flow method is used to detect dynamic feature points. Secondly, aiming at the failure of semantic segmentation network segmentation, a missed detection compensation algorithm based on the invariance of adjacent frame speed is proposed. Finally, a keyframe selection strategy is proposed to construct a semantic octree graph containing only static objects. We evaluate our algorithm on TUM RGB-D and real scene datasets. The experimental results show that the algorithm has high accuracy and real-time performance.
Currently, most works on interval valued problems mainly focus on attribute reduction (i.e., feature selection) by using rough set technologies. However, less research work on classifier building on interval-valued pr...
详细信息
Currently, most works on interval valued problems mainly focus on attribute reduction (i.e., feature selection) by using rough set technologies. However, less research work on classifier building on interval-valued problems has been conducted. It is promising to propose an approach to build classifier for interval-valued problems. In this paper, we propose a classification approach based on interval valued fuzzy rough sets. First, the concept of interval valued fuzzy granules are proposed, which is the crucial notion to build the reduction framework for the interval-valued databases. Second, the idea to keep the critical value invariant before and after reduction is selected. Third, the structure of reduction rule is completely studied by using the discernibility vector approach. After the description of rule inference system, a set of rules covering all the objects can be obtained, which is used as a rule based classifier for future classification. Finally, numerical examples are presented to illustrate feasibility and affectivity of the proposed method in the application of privacy protection.
360-degree video has shown great potential to the mainstream since its immersive experience. However, 360-degree video streaming requires ultrahigh bandwidth and low latency, which limit the improvement of user qualit...
详细信息
360-degree video has shown great potential to the mainstream since its immersive experience. However, 360-degree video streaming requires ultrahigh bandwidth and low latency, which limit the improvement of user quality of experience (QoE). Currently, methods combining field of view (FoV) prediction and adaptive video streaming provide an effective method for addressing the above issues. However, existing FoV prediction methods based on recurrent neural networks (RNN) cannot capture long-range dependency from input to output. Current deep reinforcement learning (DRL)-based adaptive strategies fail to estimate the future bandwidth with high accuracy and fully explore the capability of VR devices. To ameliorate these limitations, we design a DRL-based 360-degree video streaming method named VRFormer with FoV combined prediction and super resolution (SR). First, we adopt a content-aware transformer-based encoder-decoder network to make the long-term FoV prediction. It combines the user's head movement history, eye-tracking history, and user attention extracted from a convolutional neural network (CNN)-based network. Second, we introduce a DNN-based SR network running on VR devices to reconstruct high-definition video content. Finally, we apply a DRL-based network to adaptively allocate rates for future tiles and dynamically control video content reconstruction. Experiments have verified that the proposed method can effectively improve the quality of experience (QoE) of the user's viewing experience compared to the state-of-the-art methods.
Proportion-SVM has been deeply studied due to its broad application prospects, such as modeling voting behaviors and spam filtering. However, the geometric information has been widely ignored. Thus, current methods us...
详细信息
ISBN:
(纸本)9781509059119
Proportion-SVM has been deeply studied due to its broad application prospects, such as modeling voting behaviors and spam filtering. However, the geometric information has been widely ignored. Thus, current methods usually show sensitivity to noises. To address these problems, in this paper, we combine the proportion learning framework with Laplacian term. We exploit the advantages of Laplacian term. Extensive experiments show the effectiveness of our method.
Short text classification methods have achieved significant progress and wide application on text data such as Twitter and Weibo. However, the extremely short chinese texts like tax invoice data are different with tra...
详细信息
Short text classification methods have achieved significant progress and wide application on text data such as Twitter and Weibo. However, the extremely short chinese texts like tax invoice data are different with traditional short texts in lackness of contextual semantic information, feature sparseness and extremely short length. The existing short text classification methods are difficult to achieve a satisfactory performance in these texts. To address these problems, this paper proposes a text classification method based on bidirectional semantic extension for extremely short texts like Chinese tax invoice data. More specifically, firstly, the Chinese knowledge graph is introduced for extending bidirectional semantic of texts and label data to expand the extremely short texts and ease the problem of feature sparseness;secondly,the hash vectorization is used to avoid the semantic problem caused by the lackness of contextual information. Experimental results conducted the real tax invoice dataset demonstrate the effectiveness of our proposed method.
暂无评论