21,464 research outputs found
Towards efficient, on-demand and automated deep learning
No full textIn the past decade, deep learning has achieved great breakthroughs on tasks of computer vision, speech, language, control and many others. The advanced and dedicated computing chips, like Nvidia GPU and Google TPU, largely contributed and broadened this success. However, the requirement of large computing power impedes the deployment of deep learning methods in many real scenarios, where cost, time and energy efficiency are critical -- for example, self-driving cars, AR/VR kits, internet-of-things devices and mobile phones. This thesis presents a series of in-depth research towards efficient, on-demand and automated deep learning.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2020-08-25 without embargo termsThe student, Jiahui Yu, accepted the attached license on 2020-01-17 at 15:58.The student, Jiahui Yu, submitted this Dissertation for approval on 2020-01-17 at 16:04.This Dissertation was approved for publication on 2020-01-21 at 15:15.DSpace SAF Submission Ingestion Package generated from Vireo submission #14849 on 2020-08-25 at 17:03:08Made available in DSpace on 2020-08-26T21:53:55Z (GMT). No. of bitstreams: 3
YU-DISSERTATION-2020.pdf: 2498738 bytes, checksum: d6fc5cf2e3f6cfe94a0339c8e6a93444 (MD5)
LICENSE.txt: 4206 bytes, checksum: 9724ae490373d9c91b7c81ca6e090b30 (MD5)
PROQUEST_LICENSE.txt: 4552 bytes, checksum: f244f3fdd68510197b4e67e15280cec4 (MD5)
Previous issue date: 2020-01-2
Slimmable neural networks for edge devices
No full textWhile methods based on deep learning have witnessed major breakthroughs in machine perception and generative modeling, the problem of how to run neural networks within latency budget for edge devices remains unsolved. This thesis presents a new approach to train a single neural network executable at arbitrary widths for instant and adaptive accuracy-efficiency trade-offs at runtime.
First a simple and general method is presented to train a single neural network executable at different widths (number of channels in a layer). The width can be chosen from a predefined widths set to adaptively optimize accuracy-efficiency trade-offs at runtime. Instead of training individual networks with different width configurations, we train a shared network with switchable batch normalization. At runtime, the network can adjust its width on the fly according to on-device benchmarks and resource constraints, rather than downloading and offloading different models. Our trained networks, named slimmable neural networks, achieve ImageNet classification accuracy similar to (and in many cases better than) that of individually trained models of MobileNet v1, MobileNet v2, ShuffleNet and ResNet-50 at different widths. We also demonstrate better performance of slimmable models compared with individual ones across a wide range of applications including COCO bounding-box object detection, instance segmentation and person keypoint detection without tuning hyper-parameters. We visualize and discuss the learned features of slimmable networks.
Further, we propose a systematic approach to train universally slimmable networks (US-Nets), extending slimmable networks to execute at arbitrary width, and generalizing to networks both with and without batch normalization layers. In addition, we propose two improved training techniques for US-Nets, named the sandwich rule and the inplace distillation, to enhance training process and boost testing accuracy. We show improved performance of universally slimmable MobileNet v1 and MobileNet v2 on ImageNet classification task, compared with individually trained ones and 4-switch slimmable network baselines. We also evaluate the proposed US-Nets and improved training techniques on tasks of image super-resolution and deep reinforcement learning. Extensive ablation experiments on these representative tasks demonstrate the effectiveness of our proposed methods. Our discovery opens up the possibility to directly evaluate a FLOPs-Accuracy spectrum of network architectures. Finally, we demonstrate an application to search for channel number configurations based on proposed slimmable networks.Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2021-05-01The student, Jiahui Yu, accepted the attached license on 2019-02-14 at 14:34.The student, Jiahui Yu, submitted this Thesis for approval on 2019-02-14 at 14:42.This Thesis was approved for publication on 2019-02-15 at 11:18.DSpace SAF Submission Ingestion Package generated from Vireo submission #13390 on 2019-08-22 at 16:19:49Made available in DSpace on 2019-08-23T20:44:31Z (GMT). No. of bitstreams: 2
YU-THESIS-2019.pdf: 1268760 bytes, checksum: c091ef8a839188e9d52d208dee832b8a (MD5)
LICENSE.txt: 4206 bytes, checksum: 1b6cf1c051b15c1073c51d0ad5e1abd0 (MD5)
Previous issue date: 2019-02-15Embargo set by: Seth Robbins for item 112252
Lift date: 2021-08-23T20:44:50Z
Reason: Author requested closed access (OA after 2yrs) in Vireo ETD systemEmbargo set by: Seth Robbins for item 112252
Lift date: 2021-08-23T20:46:41Z
Reason: Author requested closed access (OA after 2yrs) in Vireo ETD systemEmbargo set by: Seth Robbins for item 112252
Lift date: 2021-08-23T20:47:38Z
Reason: Author requested closed access (OA after 2yrs) in Vireo ETD systemEmbargo set by: Seth Robbins for item 112252
Lift date: 2021-08-23T20:48:32Z
Reason: Author requested closed access (OA after 2yrs) in Vireo ETD systemLimited Restriction Lifted for Item 112252 on 2021-08-24T09:15:34Z
Jiahui Zhang, cello, Thursday, May 31, 2012
No full textIn partial fulfillment of the requirements for the degree of
Master of Musi
Hui tu zhen ben jing shi mu yu jin gang zuan
No full text著作者余好辯, 伍憤時.Cover title.上下卷.On double leaves, East Asian binding.木魚歌文.zhu zuo zhe Yu Haobian, Wu Fenshi.Shang xia juan.Mu yu ge wen
SolvBERT for solvation free energy and solubility prediction: a demonstration of an NLP model for predicting the properties of molecular complexes
No full textData and Code for "SolvBERT for solvation free energy and solubility prediction: a demonstration of an NLP model for predicting the properties of molecular complexes" by Jiahui Yu, Chengwei Zhang, Yingying Cheng, Yun-Fang Yang, Yuan-Bin She, Fengfan Liu, Weike Su, and An Su</p
sj-docx-1-pdp-10.1177_10935266231151316 – Supplemental material for Prominent Staining of MYCN Immunohistochemistry Predicts a Poor Prognosis in MYCN Non-Amplified Neuroblastoma
No full textSupplemental material, sj-docx-1-pdp-10.1177_10935266231151316 for Prominent Staining of MYCN Immunohistochemistry Predicts a Poor Prognosis in MYCN Non-Amplified Neuroblastoma by Manli Zhao, Weizhong Gu, Fei Liu, Lihua Yu, Yan Shu, Lei Liu, Jiahui Hu, Yang Liu, Hongfeng Tang and Jianhua Mao in Pediatric and Developmental Pathology</p
Yu Takeuchi
No full textYu Takeuchi is serving for JAXA since 2007 and currently working as Associate Senior Administrator at Management and Integration Department of Human Spaceflight Technology Directorate. He is also working as Researcher at the Institute of Space Law of Keio University. He received LL.M. degree from the Institute of Air and Space Law of McGill University in 2015. His main interest is in international space law inter alia the legal aspects of space traffic management and sustainable space development. He is a member of the Air Law Institute of Japan, Japanese Society of International Law, and the International Institute of Space Law (IISL).
Main Works Published in English
- “Toward the International Regime for Space Traffic Management -What to Fix the Current International Regulations-”, (November 5, 2014). Space Traffic Management Conference, Paper 23 (http://commons.erau.edu/stm/2014/wednesday/23).
- “Regulatory Regime for Tomorrow’s Suborbital Space Flights: Point-to-point International Flights”, 56th Colloquium on the Law of Outer Space, 2013.
- “Space Traffic Management as a Guiding Principle of the International Regime of Sustainable Space Activities,” 4 Journal of East Asia and International Law, 2011
- “Japanese Perspective on Legal Issues of Commercial Human Spaceflight” (co-author), 53rd Colloquium on the Law of Outer Space, 2011
- “Legal Points at Issue about NEO Threat Response and International Cooperation” (co-author), 28th International Symposium on Space Technology and Science, 2011
- “From Guideline to International Treaty for Rule of Law concerning Mitigation of Space Debris?” (co-author), 52nd Colloquium on the Law of Outer Space, 2010
Main Works Published in Japanese (title translated into English)
- “What is Space Traffic Management”, Vol. 46, No.9, Journal of the Japanese Institute of International Business Law, 2018.
- Soichiro Kozuka & Masahiko Sato eds., Introduction of Space Law for Entrepreneur (2nd. Ed.), Yuhikaku, 2018. (co-authored)
-“Challenges to International Space Law for Managing Space Traffic”, 55 Kuho (Air Law), 2014.
-“Legal Points as Issues of NEO Threat Response and International Cooperation” (co-author), 3 Spaceguard Research, Japan Spaceguard Association, 2011https://commons.erau.edu/stm-images/1121/thumbnail.jp
A Symplectic Numerical Power Flow Framework Based on Wave Finite-Element Method for Assembled Structural Systems
No full textIdentifying the propagation paths of dominant wave modes in complex assembled structure is critical for implementing wave-based vibration and noise control strategies, such as phononic band gaps. This paper presents a symplectic numerical framework to compute the wave-mode power flow in engineering assembled structures based on wave finite element method (WFEM). The power orthogonality among wave modes is explicitly formulated through the symplectic orthogonality (SO) and its adjoint form (SAO), and this formulation is further extended to the Zhong-Williams and lambda(phi) symplectic schemes. The generalized symplectic adjoint orthogonality (GSAO) and phi_SAO are subsequently proposed, providing a physically consistent basis for modal diagonalization and coherent wave propagation within the generalized symplectic eigenspace. These developments enable direct computation of the forced response and power flow entirely within the symplectic space, without reverting to the wave space. Six power-flow formulations are systematically compared and shown to yield consistent results on both beam and cylindrical shell structures. An electric motor housing is used as a case study, in which the proposed approach establishes a wave-mode power flow network. It is noted that the power-flow formulation relies on symplectic orthogonality defined for conservative WFEM systems and therefore cannot be directly applied to non-Hermitian systems
- …
