A Transfer Learning-Driven Methodology for Efficient and Sustainable Indoor Localization Using VLC
To address these challenges, we propose an energy-efficient transfer learning (TL) based approach for visible light communication based indoor localization.
ProofPaper ↗
Key points
- The rapid expansion of the Internet of Things enables seamless collaboration among connected devices, making indoor localization a critical component of industrial automation.
- However, conventional localization techniques struggle in dynamic environments.
- We also introduce a novel model efficiency (ME) metric, designed to integrate localization accuracy, energy efficiency, data efficiency, and transfer gain into a single evaluative measure for comprehensive optimization.
- These results underscore the potential of our approach to deliver a highly efficient and scalable indoor localization system suitable for energy-constrained industrial applications.
Sources (1)
- [1]A Transfer Learning-Driven Methodology for Efficient and Sustainable Indoor Localization Using VLCarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 21, 11:33 AM
To address these challenges, we propose an energy-efficient transfer learning (TL) based approach for visible light communication based indoor localization.
The rapid expansion of the Internet of Things enables seamless collaboration among connected devices, making indoor localization a critical component of industrial automation.
Extractive summary: sentences quoted from the sources.