News
- October 15th 2022: Recordings added
- September 28th 2022: Program added
- August 18th 2022: Deadline extension
- June 8th 2022: Call for papers for CODAI’22 workshop published
- May 12th 2022: CODAI has been accepted at Embedded Systems Week 2022
Overview
CODAI’22 is an Embedded Systems Week workshop to coalesce the emerging energy in the AI compiler communities and AI accelerator communities focusing on Edge AI in both academic and industrial research. These realms of research have the opportunity to deliver a pervasive and seamless end-to-end tooling that connects hardware and software development methodologies. This workshop mainly focuses on:
- Discussing new impulses for deployment of Neural Networks on low-end embedded systems
- Cooperative development of (open source) toolchains/frameworks for neural network deployment + cooperation between industry and academia
Therefore, it complements the established conferences CASES and CODES+ISSS which have a wider and more academic research-spectrum. We also welcome work-in-progress papers of ongoing research projects and case studies from industrial applications.
Topics for workshop submissions include, but are not limited to:
- Optimization techniques of neural networks on embedded edge systems: compression, pruning, quantization techniques, hardware/compiler-aware neural architecture search, etc.
- Performance estimation of neural networks on embedded systems: virtual prototyping, surrogate modeling, profiling etc.
- Compilers for Edge AI: partitioning, mapping, retargeting, intermediate representations, for heterogeneous systems and architectures, µC, DSP etc.
- Code-generation and compiler-backends for AI accelerators - especially for RISC-V and beyond-von-Neumann AI accelerators is appreciated
- Applications: processing of embedded vision, sensor signal processing, novel brain-inspired algorithms for Edge AI etc.
Program
CODAI’22 took place on October 13th 2022 as virtual workshop.
| Title | Speaker | Affiliation | Video | Slides |
|---|---|---|---|---|
| Introduction | Michael J. Klaiber | CODAI General Co-Chair, EnCharge AI | watch | |
| Industrial Keynote: The Past, Presence and Future of AI Compute Platforms | Kailash Gopalakrishnan | EnCharge AI | watch | |
| Performance models and energy-optimal mapping of DNNs on many-core hardware with dynamic power management | Bernhard Vogginger, Florian Kelber, Johannes Partzsch and Christian Mayr | TU Dresden | watch | slides |
| HyT-NAS: Hybrid Transformers Neural Architecture Search for Edge Devices | Lotfi Abdelkrim Mecharbat, Hadjer Benmeziane, Ouarnoughi Hamza and Smail Niar | Ecole Nationale Supérieure d’Informatique, Hauts-de-France, Université Polytechnique Hauts-de-France | watch | slides |
| Whole-model optimization with Apache TVM | Andrew Reusch | OctoML | watch | |
| MLonMCU: TinyML Benchmarking with Fast Retargeting | Philipp van Kempen, Rafael Stahl, Daniel Müller-Gritschneder and Ulf Schlichtmann | Technical University of Munich | watch | slides |
| Academic Keynote: Abstract for Machine Learning Compilations | Tianqi Chen | Carnegie Mellon University | watch | |
| Embedded AI Challenges in Perception Systems | Miguel Aguilar | Aptiv | watch | |
| Integration of a systolic array based hardware accelerator into a DNN operator auto-tuning framework | Federico Nicolás Peccia and Oliver Bringmann | FZI Research Center for Information Technology, University of Tübingen | watch | slides |
| Deploying Machine Learning Models to Ahead-of-Time Runtime on Edge Using MicroTVM | Chen Liu, Liyuan Guo, Xinyue Shi, Matthias Jobst, Johannes Partzsch and Christian Mayr | Chair of Highly-Parallel VLSI-Systems and Neuromorphic Circuits, TU Dresden | watch | |
| Conclusion | Sebastian Vogel | CODAI General Co-Chair, NXP | watch |
Important Dates
- Abstract submission deadline: August 25th 2022 AOE
- Paper submission deadline: September 8th 2022 AOE
August 18th 2022 AOE - Acceptance notification: September 26th, 2022
September 22nd, 2022September 8th, 2022 - Camera-ready paper: October 7th, 2022
September 30th, 2022 - Workshop date: October 13th 2022
Call For Papers
Call for papers for CODAI’22 workshop
Paper Submission
Registration
ESWEEK Registration $15.75 / $20.00
If you have registered the main conference, you can attend the workshop for free.
Author Instructions
- Submit camera-ready via Easychair by October 7th, 2022
- Use IEEE Template
- Paper must not exceed a length of 5 pages incl. references
- Submit slides as PPT or PDF by October 12th 2022
- via Github pull request here
- add link to index.md
Organization
Organization Committee
- Michael J. Klaiber - EnCharge AI
- Sebastian Vogel - NXP Semiconductors, The Netherlands
Program Committee
- Deming Chen - University of Illinois Urbana Champaign, USA
- Andreas Bytyn - Bosch Automotive Electronics, Germany
- Andrew Reusch - OctoML, USA
- Henk Corporaal - Eindhoven University of Technology, The Netherlands
- Robert Nabholz - RB XC, Germany
- Floran de Putter - Eindhoven University of Technology, The Netherlands
- Oliver Bringmann - University of Tübingen, Germany
- Paul Palomero Bernardo - University of Tübingen, Germany
- Falk Rehm - Bosch Research, Germany
- Dennis Rieber - Bosch Research, Germany
- Christoph Schorn - Bosch Research, Germany
- Ingo Feldner - Bosch Research, Germany
- Hussam Amrouch - University of Stuttgart, Germany
- Dayane Reis, University of South Florida, USA
Contact
For questions contact:
Dr. Michael J. Klaiber, EnCharge AI
E-Mail: codai2022 (a) michael-klaiber.de
Logo by Parzival’ 1997