Table 3 Configuration of the edge nodes and the cloud servers.
From: Leveraging federated learning and edge computing for pandemic-resilient healthcare
Sl. No | Node | Description |
|---|---|---|
1 | Raspberry Pi 4 (Edge) | DNN Model, YOLO v4, Programming based on Python, RAM in GB- 4GB, Processor is ARM Cortex-72, Clock- 1.5 GHz, GPU Pi3 A+: Broadcom, VideoCore 1V; 400 MHz |
2 | Grove AI Hat (Edge) | DNN model K-model, Programming based on MicroPython RAM in GB 0.008, Processor M1K210 RISC-V Clock in GHz 0.04 to 0.06 GPU KPU Maximum training time 120 min Response time after training 2 s |
3 | Google Colab | DNN model RESNET-50, YOLOv4, MobileNet, VGG-16, SocialDistanceNet-19 etc, Programming based on Python, RAM in GB 32, Processor Intel ®Xenon ®Gold5318N, Clock in GHz 2.25, GPU Tesla T4 |
4 | FL Platform | DNN model YOLOv4+SENET, Node used: 2, 3, 4, 6, 8, 10, Optimize- ADAM, Cost Function-Sparse Categorial |
