Exploring 2D representation and transfer learning techniques for indoor localization.
Saved in:
| Title: | Exploring 2D representation and transfer learning techniques for indoor localization. |
|---|---|
| Authors: | Kerdjidj, O.1,2 (AUTHOR) kerdjidjoussama@gmail.com, Himeur, Y.2 (AUTHOR), Atalla, S.2 (AUTHOR), Copiaco, A.2 (AUTHOR), Sohail, S S.3 (AUTHOR), Amira, A.4,5 (AUTHOR), Fadli, F.6 (AUTHOR), Mansoor, W.2 (AUTHOR), Gawanmeh, A.2 (AUTHOR) |
| Source: | Multimedia Tools & Applications. Oct2025, Vol. 84 Issue 33, p40707-40719. 13p. |
| Subjects: | Deep learning, Two-dimensional models, Indoor positioning systems, Feature extraction, Wavelets (Mathematics), Transfer of training |
| Abstract: | Indoor localization (IL) plays a critical role in various applications, including asset tracking, emergency response, and indoor navigation. Traditional methods relying on one-dimensional (1D) representations of received signal strength indicator (RSSI) data often suffer from limitations such as low accuracy, environmental sensitivity, and susceptibility to interference. To overcome these challenges, this paper proposes Scalogram-Based Transfer Learning for Indoor Localization (STL-IL), a novel framework that significantly enhances localization performance. The proposed approach transforms 1D RSSI signals into two-dimensional (2D) representations using scalogram-based time-frequency analysis, coupled with advanced transfer learning (TL) techniques. By leveraging the scalogram method to generate image representations of the data and employing diverse wavelet types. Furthermore, pre-trained deep learning models, including GoogleNet and SqueezeNet, are utilized to achieve state-of-the-art performance. Experimental results demonstrate that the STL-IL system achieves remarkable accuracy levels of 95% and 96% for SqueezeNet and GoogleNet, respectively. The proposed framework not only improves localization accuracy but also enhances robustness in dynamic and challenging environments, making it a promising solution for real-world indoor localization applications. [ABSTRACT FROM AUTHOR] |
| Copyright of Multimedia Tools & Applications is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | Indoor localization (IL) plays a critical role in various applications, including asset tracking, emergency response, and indoor navigation. Traditional methods relying on one-dimensional (1D) representations of received signal strength indicator (RSSI) data often suffer from limitations such as low accuracy, environmental sensitivity, and susceptibility to interference. To overcome these challenges, this paper proposes Scalogram-Based Transfer Learning for Indoor Localization (STL-IL), a novel framework that significantly enhances localization performance. The proposed approach transforms 1D RSSI signals into two-dimensional (2D) representations using scalogram-based time-frequency analysis, coupled with advanced transfer learning (TL) techniques. By leveraging the scalogram method to generate image representations of the data and employing diverse wavelet types. Furthermore, pre-trained deep learning models, including GoogleNet and SqueezeNet, are utilized to achieve state-of-the-art performance. Experimental results demonstrate that the STL-IL system achieves remarkable accuracy levels of 95% and 96% for SqueezeNet and GoogleNet, respectively. The proposed framework not only improves localization accuracy but also enhances robustness in dynamic and challenging environments, making it a promising solution for real-world indoor localization applications. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 13807501 |
| DOI: | 10.1007/s11042-025-20780-8 |
