We are entering an era full of modern AI-powered mobile/IoT applications with prominent examples like virtual/augmented reality and autonomous driving. Leveraging sophisticated machine learning technologies, these applications typically require massive computation applied on an enormous amount of data (e.g., audio/video and time-series sensor data). On the other hand, such applications impose stringent constraints on performance metrics like latency, throughput, and reliability due to their user-interactive and mission-critical nature. These challenges have driven the evolution of the traditionally centralized cloud computing paradigm into a more distributed one at the network edge in close proximity of end-devices or even on-device computing solutions, where energy efficiency is of critical concern.
WEEE focuses on efficient machine learning solutions for mobile/IoT scenarios and energy-efficient systems for edge computing and networking in general. We invite researchers and practitioners to submit original ongoing work, visionary ideas, or experience related to topics include, but not limited to:
Yunxin Liu, Tsinghua University
Title: On-Device Deep Learning: from Centralized Intelligence in the Cloud to Distributed Intelligence on the Edge
Abstract: With the advances of hardware, software, and artificial intelligence (AI), there is a new computing paradigm shift from centralized intelligence in the cloud to distributed intelligence on the edge. In the era of edge computing, it is critical to infuse AI to empower diverse edge devices and applications. This talk overviews the challenges and opportunities of on-device deep learning; and introduces our recent research work on making on-device deep-learning more efficient, focusing on how to build affordable AI models customized for diverse edge devices, and how to maximize the performance of on-device model inference by fully utilizing the heterogeneous computing resources.
Bio: Yunxin Liu is a Guoqiang Professor and Principal Investigator at Institute for AI Industry Research (AIR), Tsinghua University. He joined Tsinghua University in April 2021, prior to which he was a Principal Research Manager at Microsoft Research Asia. His research interests are mobile and edge computing. His research work has been published in top conferences and journals such as MobiSys, MobiCom, SenSys, NSDI, CCS, TON, TMC and TPDS; transferred into multiple Microsoft products including Visual Studio, XBOX XDK and Windows Phone; and featured in news media worldwide. He received MobiCom 2015 Best Demo Award, PhoneSense 2011 Best Paper Award, and SenSys 2018 Best Paper Runner-up Award. He is a General Co-chair of MobiHoc 2021 and an Associate Editor of IEEE Transactions on Mobile Computing.
All time in Central European Summer Time (CEST), UTC+2.
|14:00 - 14:10||Opening by the organizers|
|14:10 - 15:00||Keynote by Yunxin Liu (Tsinghua University)|
|15:00 - 15:10||Break|
|15:10 - 16:25||
Session 1 (3 paper presentations, 25 min each)
- Design Considerations for Energy-efficient Inference on Edge Devices
- Efficiency of Virtualization over MEC plus Cloud
- Energy-efficient AI over a Virtualized Cloud Fog Network
|16:25 - 16:35||Break|
|16:35 - 17:25||
Session 2 (2 paper presentations, 25 min each)
- Automated Selection of Energy-efficient Operating System Configurations
- Self-Sustainable Cyber-Physical Systems with Collaborative Intermittent Computing
|17:25 - 17:30||Closing|
Design Considerations for Energy-efficient Inference on Edge Devices
Walid A. Hanafy, Tergel Molom-Ochir, Rohan Shenoy (University of Massachusetts, Amherst, USA)
Automated Selection of Energy-efficient Operating System Configurations
Benedict Herzog, Fabian Hügel, Stefan Reif, Wolfgang Schröder-Preikschat (Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany), Timo Hönig (Ruhr-Universität Bochum, Germany)
Self-Sustainable Cyber-Physical Systems with Collaborative Intermittent Computing
Gaosheng Liu, Lin Wang (Vrije Universiteit Amsterdam, The Netherlands)
Efficiency of Virtualization over MEC plus Cloud (invited)
Vincenzo Mancuso (IMDEA Networks Institute, Spain), Paolo Castagno, Matteo Sereno (Università di Torino, Italy)
Energy-efficient AI over a Virtualized Cloud Fog Network (invited)
Barzan A. Yosuf, Sanaa H. Mohamed, Mohammed M. Alenazi, Taisir E. H. El-Gorashi, Jaafar M. H. Elmirghani (University of Leeds, UK)
Submissions focusing on specific mobile/IoT applications and generic edge computing are both welcome. We solicit the following two types of contributions:
- Regular papers describing original research ideas and work, up to 6 pages including all figures, tables, but excluding references. Accepted regular papers will be presented at the workshop as oral presentations.
- Vision papers identifying new research problems/challenges in the field, up to 4 pages including all figures, tables, but excluding references. Accepted vision papers will be presented at the workshop as short presentations followed by interactive discussions.
Papers that, at the time of submission, are under review for another venue or have been published, or that overlap substantially in technical content with papers under review or previously published, should not be submitted. All accepted regular and vision papers will be published in the conference proceedings and the ACM Digital Library.
We only accept submissions in the PDF format. All submissions should be formatted using the standard 9-point ACM double-column format (sigconf proceedings template), single-blind. Papers that do not meet the size and formatting requirements may not be reviewed. Word and LaTeX templates are available on the ACM Publications Website. Submissions will be judged on originality, relevance, technical quality, and clarity. At least one of the authors of every accepted paper must register and present the work at the workshop.
Submission site: https://weee2021.hotcrp.com.
- Submission deadline:
Mar 31, 2021April 14, 2021 (extended)
- Author notification: Apr 30, 2021
- Camera-ready submission: May 18, 2021
@ 2021 WEEE Organizing Committee. All rights reserved.