Institution
Huawei
Company•Shenzhen, China•
About: Huawei is a company organization based out in Shenzhen, China. It is known for research contribution in the topics: Terminal (electronics) & Node (networking). The organization has 41417 authors who have published 44698 publications receiving 343496 citations. The organization is also known as: Huawei Technologies & Huawei Technologies Co., Ltd..
Papers published on a yearly basis
Papers
More filters
•
25 Sep 2013TL;DR: In this article, the authors present a method and a device for clearing malicious power consumption applications, and a user terminal Background work power consumption of applications running in the user terminal is periodically counted.
Abstract: An embodiment of the invention provides a method and a device for clearing malicious power consumption applications, and a user terminal Background work power consumption of applications running in the user terminal is periodically counted, the applications with the background work power consumption not lower than a power consumption threshold value are determined as the malicious power consumption applications, wake lock holding time of each application running in the user terminal with a screen closed is periodically counted, if a certain application with the holding time not shorter than a set time threshold value is a background work application, the application is determined as a malicious power consumption application unreasonably occupying resources in background, the applications with high power consumption but normally used by a user are not malicious power consumption applications, the malicious power consumption applications can be accurately positioned and detected, unnecessary power consumption of the user terminal is avoided while usage experience of the user is ensured, electric energy is saved, and the battery life of the user terminal is prolonged to a certain degree
63 citations
•
07 Aug 2008TL;DR: In this paper, a mobile station device transmits a preamble to a base station and performs uplink timing alignment based on the synchronization timing deviation information included in a random access response.
Abstract: A mobile station device transmits a random access preamble to a base station device and performs uplink timing alignment based on the synchronization timing deviation information included in a random access response which the base station device transmits in response to the transmitted random access preamble, wherein in an uplink synchronous status, the mobile station device does not perform uplink timing alignment based on synchronization timing deviation information included in a random access response, which is a response to a random access preamble whose preamble ID is randomly selected by the mobile station device.
63 citations
••
13 Aug 2017TL;DR: This paper presents a new system called STREAMDM-C++, that implements decision trees for data streams in C++, and that has been used extensively at Huawei, and is easy to extend, and contains more powerful ensemble methods, and a more efficient and easy to use adaptive decision trees.
Abstract: Nowadays real-time industrial applications are generating a huge amount of data continuously every day. To process these large data streams, we need fast and efficient methodologies and systems. A useful feature desired for data scientists and analysts is to have easy to visualize and understand machine learning models. Decision trees are preferred in many real-time applications for this reason, and also, because combined in an ensemble, they are one of the most powerful methods in machine learning. In this paper, we present a new system called STREAMDM-C++, that implements decision trees for data streams in C++, and that has been used extensively at Huawei. Streaming decision trees adapt to changes on streams, a huge advantage since standard decision trees are built using a snapshot of data, and can not evolve over time. STREAMDM-C++ is easy to extend, and contains more powerful ensemble methods, and a more efficient and easy to use adaptive decision trees. We compare our new implementation with VFML, the current state of the art implementation in C, and show how our new system outperforms VFML in speed using less resources.
63 citations
••
TL;DR: In this article, the authors provide technical insight and rationales behind the recently approved ITU-T G.989.2 Recommendation of the 40-gigabit-capable passive optical networks (NG-PON2).
Abstract: This is the second of a two-part paper intended to provide technical insight and rationales behind the recently approved ITU-T G.989.2 Recommendation: the physical media dependent layer specification of the 40-gigabit-capable passive optical networks (NG-PON2). While Part 1 of the paper discusses topics related to the optical link design, Part 2 focuses on wavelength control, technology feasibility, management and control channel design, and potential future standardization directions of such a multi-wavelength PON system. As the NG-PON2 system will continue to evolve, technology extensions are also discussed to provide guidance for future research.
63 citations
•
TL;DR: KSE is capable of simultaneously compressing each layer in an efficient way, which is significantly faster compared to previous data-driven feature map pruning methods, and significantly outperforms state-of-the-art methods.
Abstract: Compressing convolutional neural networks (CNNs) has received ever-increasing research focus. However, most existing CNN compression methods do not interpret their inherent structures to distinguish the implicit redundancy. In this paper, we investigate the problem of CNN compression from a novel interpretable perspective. The relationship between the input feature maps and 2D kernels is revealed in a theoretical framework, based on which a kernel sparsity and entropy (KSE) indicator is proposed to quantitate the feature map importance in a feature-agnostic manner to guide model compression. Kernel clustering is further conducted based on the KSE indicator to accomplish high-precision CNN compression. KSE is capable of simultaneously compressing each layer in an efficient way, which is significantly faster compared to previous data-driven feature map pruning methods. We comprehensively evaluate the compression and speedup of the proposed method on CIFAR-10, SVHN and ImageNet 2012. Our method demonstrates superior performance gains over previous ones. In particular, it achieves 4.7 \times FLOPs reduction and 2.9 \times compression on ResNet-50 with only a Top-5 accuracy drop of 0.35% on ImageNet 2012, which significantly outperforms state-of-the-art methods.
63 citations
Authors
Showing all 41483 results
Name | H-index | Papers | Citations |
---|---|---|---|
Yu Huang | 136 | 1492 | 89209 |
Xiaoou Tang | 132 | 553 | 94555 |
Xiaogang Wang | 128 | 452 | 73740 |
Shaobin Wang | 126 | 872 | 52463 |
Qiang Yang | 112 | 1117 | 71540 |
Wei Lu | 111 | 1973 | 61911 |
Xuemin Shen | 106 | 1221 | 44959 |
Li Chen | 105 | 1732 | 55996 |
Lajos Hanzo | 101 | 2040 | 54380 |
Luca Benini | 101 | 1453 | 47862 |
Lei Liu | 98 | 2041 | 51163 |
Tao Wang | 97 | 2720 | 55280 |
Mohamed-Slim Alouini | 96 | 1788 | 62290 |
Qi Tian | 96 | 1030 | 41010 |
Merouane Debbah | 96 | 652 | 41140 |