Progress in the field of Human Activity Recognition (HAR) has taken a big leap forward thanks to the scientific paper ‘Efficient Tiny Machine Learning for Human Activity Recognition on Low-Power Edge Devices’, written by Vinamra Sharma, Danilo Pietro Pau and José Cano.
This work, part of the dAIEDGE project, has been awarded with a prestigious Track Paper Award, standing out as a reference in the field of Artificial Intelligence and machine learning on low-power edge devices.
The paper introduces methods that achieve remarkable advances in the optimisation of TinyML models. Some of the most outstanding results include:
- Reduction of the model footprint by an average of 4×.
- Nearly intact accuracy, with only 4% deviation.
- Unprecedented energy efficiency.
Where can you read this article?
If you want to know more details about this exciting breakthrough in the field of TinyML and HAR, you can access the full article here.