Fall-Detection-with-Federated-Transfer-Learning

Leveraging Federated Learning and Transfer Learning for identifying FALL and ADL from images. We have employed:

• Feature Extractor: VGG16
• Aggregation Technique: FedAvg

This work has been published as a conference proceedings in BIM 2023, titled Federated Transfer Learning for Vision-Based Fall Detection.

---

Dataset

UR Fall Detection Dataset -- It contains 70 videos illustrating FALL and ADL events. It also provides RGB and Depth data for each video. We've utilized the RGB frames for this work and augmented the dataset by employing common augmentation techniques (e.g. rotation, flipping, etc.).

Citation

@InProceedings{10.1007/978-981-99-8937-9_64,
author="Mistry, Durjoy
and Tonmoy, Moshiur Rahman
and Anower, Md. Shahib
and Hasan, A. S. M. Touhidul",
title="Federated Transfer Learning for Vision-Based Fall Detection",
booktitle="Proceedings of the 2nd International Conference on Big Data, IoT and Machine Learning",
year="2024",
publisher="Springer Nature Singapore",
address="Singapore",
pages="961--975"
}