Tiny Machine Learning

Special session @ WCCI 2022

Paper submission: January 31, 2022

What is tiny machine learning?

Tiny Machine Learning is a novel research area aiming at designing and developing Machine and Deep Learning techniques meant to be executed on Embedded Systems, Internet-of-Things and Edge Computing units, hence taking into account the constraints on computation, memory, and energy characterizing these pervasive devices.

Techniques to significantly reduce the size and power of models can draw from a vast number of training methods, ranging from pruning to dynamic compression and quantization, to the design of novel hardware accelerators or distributed layers and algorithms specific for execution on “tiny” platforms.

This special issue aims to bring together these trends, providing a unified view of the research and applications concerning tiny scenarios.

Submitting a paper

Submission is done through the WCCI 2022 website, by selecting "Tiny Machine Learning" as special session when submitting.

All requirements from the main conference apply, in particular:

  1. The review process for WCCI 2022 will be double-blind;

  2. Each paper is limited to 8 pages, including figures, tables, and references;

  3. The review process for WCCI 202WCCI 2022 adopts Microsoft CMT as submission system, available at the following link: https://cmt3.research.microsoft.com/IEEEWCCI2022/

Paper submission is January 31, 2022 (11:59 PM AoE), Notification of acceptance: April 26, 2022, Final paper submission: May 23, 2022.


Papers must present original work or review the state-of-the-art in the following non-exhaustive list of topics:

  • Approximate computing mechanisms

  • Distributed Tiny Machine Learning and Inference

  • Custom hardware for Tiny Machine Learning

  • On-device learning for Tiny Machine Learning

  • Smart objects and IoT devices

  • Low-power Machine and Deep Learning algorithms

  • Intelligence for embedded systems

  • Computational intelligence for cyber-physical systems

  • Intelligent fault diagnosis systems

  • Adaptive solutions to operate in evolving/changing environments

  • Intelligent systems for real-world applications

This special session aspires at building a bridge between academic and industrial research, as well as among researchers working in different fields, with the specific purpose of designing IoT and embedded systems endowed with machine and deep learning capabilities able to adapt and interact with evolving environments.

Special session organizers

Manuel Roveri, Politecnico di Milano

Simone Scardapane, Sapienza University of Rome

Torsten Hoefler, ETH, Switzerland