Organized by the Educational Board of the IEEE Signal Processing Society

The Educational Board invites proposals for short, hands-on courses (~3 hours) tailored for industry professionals. Selected courses will be featured in our 2026 professional development series and made available at IEEE Learning Network. You provide the content; IEEE will handle the production.

 Focus Areas of Interest (but topics not listed are encouraged!)

1. Foundation Models in Industry
   This course focuses on utilizing multimodal LLMs (text, audio and video) for reasoning and generation. It explores how industry practitioners can deploy these models as the "brains" of physical agentic and robotic systems. The curriculum covers integrating continuous signal streams with LLMs, leveraging VLA architectures, and deploying signal-based RAG to enable large models to interact with physical environments and real-world sensors.
2. Edge AI: Cost-Effective Inference and Efficient Signal Processing
   Addressing the growing industrial interest in edge computing and embedded systems, this course focuses on maximizing inference efficiency and minimizing deployment costs. To achieve highly cost-effective ML inference, the course covers modern deployment strategies, including extreme model quantization, hardware-aware optimization, neuromorphic engineering, and on-device execution on low-cost embedded hardware.
3. Trustworthy AI and Signal Forensics 
   Gaining importance across many industries, this area is framed around tangible industrial threats and solutions. The course focuses on audio/video deepfake detection, signal watermarking, adversarial robustness in sensor data, and secure perception pipelines for autonomous systems.
4. Physics-Informed and Signal-Driven Machine Learning
   Integrating signal processing into data-driven, black-box machine learning models. Industry struggles with the "black-box" nature of pure deep learning (hallucinations, lack of explainability, poor out-of-distribution generalization). This course is about how injecting DSP priors into architectures like YOLOv11 or Mamba improves data efficiency, interpretability, and robust performance in real-world environments.

Timeline

• Submission Deadline: June 30th, 2026
• Course Delivery: Throughout 2026

Submit proposals to:        https://forms.gle/q8nLyQqa1v5EeQuL7 

Contribute to the advancement of signal processing education for working professionals.

Submit your course proposal and help shape the future of industry-focused learning!

Organizing committee

• Milos Cernak, milos.cernak@ieee.org
• Hala As'ad, hala.asad@ieee.org 
• Vicky Lu, vicky.h.lu@gmail.com