Applications, Architectures, Methods and Tools for Machine - and Deep Learning (AAMTM)

Machine learning has numerous important applications in intelligent systems within many areas, like automotive, avionics, robotics, health-care, well-being, and security. The recent progress in Artificial Intelligence (AI), and particularly in Deep Learning (DL) / Machine Learning (ML), has dramatically improved the state-of-the-art in object detection, classification and recognition, natural language processing, games, medical imaging, etc. However, the complexity of DL-networks for many practical applications can be huge, and their processing may demand a high computing effort and excessive energy consumption. This can become a gigantic challenge when considering embedded inference implementation for Smart Cyber Physical Systems (like autonomous vehicles and robotics) and Internet-of-Things (like healthcare-IoT and predictive maintenance for Industry 4.0). Moreover, even training of such complex DL-networks over massive data sets is triggering new avenues in training accelerator design. In DSD 2020, we plan to organize several oral sessions on embedded deep learning/AI and related research, as well as to have invited speeches, and a poster session.

Special Session Scope

We welcome submissions related to advanced applications, architectures, design methods and tools, and system software for AI, ML and DL, especially related (but not limited) to the following topics:

• Architectures for ML and DL, with emphasis on energy reduction, computation efficiency and/or computation flexibility, both for inference and/or for learning
• Neuromorphic architectures, Spiking and brain-inspired neural networks and their implementation
• Efficient mapping of ML and DL applications to target architectures, including many-core, GPGPU, SIMD, FPGA, and HW accelerators
• New learning approaches for ML and DL, with emphasis on e.g. faster and more efficient learning, online learning, and quality of learning, training accelerators, etc.
• High-level programming language support for ML and DL
• Advanced applications exploiting ML or DL
• ML and DL for design automation
• Tools, frameworks, and system software for ML and DL
• Using of approximate computing to decrease the energy demands of ML and DL
• Security and Reliability issues for ML and DL, for both inference and training

Submission Guidelines

Authors are encouraged to submit their manuscripts via EasyChair web service at web page https://easychair.org/conferences/?conf=dsd2021. Each manuscript should include the complete paper text, all illustrations, and references. The manuscript should conform to the IEEE format: single-spaced, double column, US letter page size, 10-point size Times Roman font, up to 8 pages. In order to conduct a blind review, no indication of the authors' names should appear in the manuscript, references included.
CPS, Conference Publishing Services, publishes the (ISI indexed) DSD Proceedings, available worldwide through the IEEE Xplore Digital Library. Extended versions of selected best papers will be published in a special issue of the ISI indexed “Microprocessors and Microsystems: Embedded Hardware Design” Elsevier journal.

Special Session Chair

M. Shafique (TU Wien, A)

Special Session Program Committee

Ghayoor Abbas Gillani, University of Twente

Muhammad Shafique, Vienna University of Technology

Satwik Patnaik, Texas A&M University

Lilas Alrahis, NYU Abu Dhabi

Emanuele Torti, University of Pavia

Ihsen Alouani, IEMN-DOAE/UMR CNRS, Polytechnic University Hauts-de-France

Alberto Marchisio, Vienna University of Technology

Marco La Salvia, University of Pavia

Vojtech Mrazek, Brno University of Technology

Marcelo Brandalero, Brandenburg University of Technology Cottbus-Senftenberg

Contact Information

Prof. Muhammad Shafique
e-mail: muhammad.shafique@tuwien.ac.at