You are required to read and agree to the below before accessing a full-text version of an article in the IDE article repository.
The full-text document you are about to access is subject to national and international copyright laws. In most cases (but not necessarily all) the consequence is that personal use is allowed given that the copyright owner is duly acknowledged and respected. All other use (typically) require an explicit permission (often in writing) by the copyright owner.
For the reports in this repository we specifically note that
- the use of articles under IEEE copyright is governed by the IEEE copyright policy (available at http://www.ieee.org/web/publications/rights/copyrightpolicy.html)
- the use of articles under ACM copyright is governed by the ACM copyright policy (available at http://www.acm.org/pubs/copyright_policy/)
ss - technical reports and other articles issued by Mälardalen University is free for personal use. For other use, the explicit consent of the authors is required
- in other cases, please contact the copyright owner for detailed information
By accepting I agree to acknowledge and respect the rights of the copyright owner of the document I am about to access.
If you are in doubt, feel free to contact webmaster@ide.mdh.se
DPAC seminar - Mohammad Loni
Type:
Start time:
End time:
Location:
Description
Title: DeepMaker: Customizing Deep Neural Networks for COTS Programmable Devices
Abstract: Cyber Physical Systems (CPS) have become immensely prominent everywhere due to their considerable benefits including improving safety and reliability, providing higher quality of service and human’s convenience. Expanding the applicability of CFS in different aspects of life causes to emerge genesis new classes of applications such as recognition, mining, analytics, and search. Deep Machine Learning (ML) algorithms play a key role in these applications by providing high estimation accuracy. However we need to capture and process a big amount of raw data for modern applications such as autonomous system, and smart factories. Moreover, for enabling tremendous improvements in achieving accuracy, deep learning models moved toward emerging new deeper and deeper complex network architectures. Although using high-performance processing infrastructures like cloud data centers are prevalent, we aim to find a near-sensor processing solution since in some CPS we are interested in keeping processing close to sensors due to security reasons, limited network bandwidth, increasing the energy efficiency and guaranteeing worse case execution time. To reach this goal, we offer a comprehensive framework, named DeepMaker, aiming to generate a highly robust DNN accelerator for both embedded devices and high-performance cloud infrastructures. In nutshell, DeepMaker tries to bring processing close to embedded devices as much as possible.