RT3 Publications: Edge Computing for AI
2024
Al-Ameen, Shamil; Sudharsan, Bharath; Vijayakumar, Tejus; Szydlo, Tomasz; Shah, Tejal; Ranjan, Rajiv
Poly Instance Recurrent Neural Network for Real-time Lifelong Learning at the Low-power Edge Proceedings Article
In: 2024 IEEE International Conference on Big Data (BigData), pp. 5907-5916, 2024.
@inproceedings{10825026,
title = {Poly Instance Recurrent Neural Network for Real-time Lifelong Learning at the Low-power Edge},
author = {Shamil Al-Ameen and Bharath Sudharsan and Tejus Vijayakumar and Tomasz Szydlo and Tejal Shah and Rajiv Ranjan},
doi = {10.1109/BigData62323.2024.10825026},
year = {2024},
date = {2024-01-01},
urldate = {2024-01-01},
booktitle = {2024 IEEE International Conference on Big Data (BigData)},
pages = {5907-5916},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
2023
Al-Ameen, Shamil; Sudharsan, Bharath; Szydlo, Tomasz; Al-Taie, Roua; Shah, Tejal; Ranjan, Rajiv
Tiny-Impute: A Framework for On-device Data Quality Validation, Hybrid Anomaly Detection, and Data Imputation at the Edge Bachelor Thesis
2023.
@bachelorthesis{Al-Ameen2023,
title = {Tiny-Impute: A Framework for On-device Data Quality Validation, Hybrid Anomaly Detection, and Data Imputation at the Edge},
author = {Shamil Al-Ameen and Bharath Sudharsan and Tomasz Szydlo and Roua Al-Taie and Tejal Shah and Rajiv Ranjan},
doi = {10.1145/3603166.3632164},
year = {2023},
date = {2023-12-04},
urldate = {2023-12-04},
pages = {1–10},
publisher = {ACM},
keywords = {},
pubstate = {published},
tppubtype = {bachelorthesis}
}
2022
Sun, Rui; Li, Yinhao; Shah, Tejal; Sham, Ringo W. H.; Szydlo, Tomasz; Qian, Bin; Thakker, Dhaval; Ranjan, Rajiv
FedMSA: A Model Selection and Adaptation System for Federated Learning Journal Article
In: Sensors, vol. 22, no. 19, 2022, ISSN: 1424-8220.
@article{Sun2022,
title = {FedMSA: A Model Selection and Adaptation System for Federated Learning},
author = {Rui Sun and Yinhao Li and Tejal Shah and Ringo W. H. Sham and Tomasz Szydlo and Bin Qian and Dhaval Thakker and Rajiv Ranjan},
doi = {10.3390/s22197244},
issn = {1424-8220},
year = {2022},
date = {2022-10-00},
urldate = {2022-10-00},
journal = {Sensors},
volume = {22},
number = {19},
publisher = {MDPI AG},
abstract = {<jats:p>Federated Learning (FL) enables multiple clients to train a shared model collaboratively without sharing any personal data. However, selecting a model and adapting it quickly to meet user expectations in a large-scale FL application with heterogeneous devices is challenging. In this paper, we propose a model selection and adaptation system for Federated Learning (FedMSA), which includes a hardware-aware model selection algorithm that trades-off model training efficiency and model performance base on FL developers’ expectation. Meanwhile, considering the expected model should be achieved by dynamic model adaptation, FedMSA supports full automation in building and deployment of the FL task to different hardware at scale. Experiments on benchmark and real-world datasets demonstrate the effectiveness of the model selection algorithm of FedMSA in real devices (e.g., Raspberry Pi and Jetson nano).</jats:p>},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
<jats:p>Federated Learning (FL) enables multiple clients to train a shared model collaboratively without sharing any personal data. However, selecting a model and adapting it quickly to meet user expectations in a large-scale FL application with heterogeneous devices is challenging. In this paper, we propose a model selection and adaptation system for Federated Learning (FedMSA), which includes a hardware-aware model selection algorithm that trades-off model training efficiency and model performance base on FL developers’ expectation. Meanwhile, considering the expected model should be achieved by dynamic model adaptation, FedMSA supports full automation in building and deployment of the FL task to different hardware at scale. Experiments on benchmark and real-world datasets demonstrate the effectiveness of the model selection algorithm of FedMSA in real devices (e.g., Raspberry Pi and Jetson nano).</jats:p>