Keynote on Efficient Human Activity Recognition at the IWOAR Workshop

Dr. Özlem Durmaz-Incel delivered a keynote speech at the 9th International Workshop on Sensor-Based Activity Recognition and Artificial Intelligence (IWOAR) held at HPI in Potsdam.

Titled "Beyond the Basics: Enhancing Lightweight Models for Resource-efficient Human Activity Recognition," Dr. Durmaz-Incel's talk explored methods for improving the performance of lightweight models designed for sensor-based human activity recognition (HAR) on resource-constrained devices, specifically wearables with limited memory. Her presentation highlighted the use of convolutional block attention modules and knowledge distillation techniques to achieve high recognition accuracy while maintaining computational efficiency.

Learn More:

• Workshop webpage: https://iwoar.org/2024/