Program

Part 1: Algorithms [45 min] [Radu Marculescu slides]

• 1.1 Edge AI (motivation/challenges)

• 1.2 Federated learning

• 1.3 Deep learning model compression

Part 2: Architectures [45 min] [Diana Marculescu slides ]

• 2.1 System-aware modeling

• 2.2 ML and system co-design via NAS

• 2.3 Efficient interference at the edge

Break [30 min]

Part 3: Applications [45 min] [ Umit Ogras slides ]

• 3.1 Sensor-rich energy-harvesting devices

• 3.2 Human activity recognition

• 3.3 Gait analysis using wearables

Part 4: Demos/Panel/Q&A [45 min] [All live]

We will play our tutorial videos on Zoom starting 1:00pm PST, so please join us on time. We will answer questions from the audience in-real time in the Zoom chat. During Part 4, we will host a live panel and an extended Q&A session.