Federated-Benchmark: A Benchmark of Real-world Images Dataset for Federated Learning

Overview

We present a real-world image dataset, reflecting the characteristic real-world federated learning scenarios, and provide an extensive benchmark on model performance, efficiency, and communication in a federated learning setting.

Get Start

• Please firstly ensure you have CUDA 10.2 installed.
• Run make to install dependencies, build cython files, and generate task json files.
• For training, simply run make run to start and make stop to stop the training process.
  • during training, use make watch to see if the training is going on;
  • use make run NUM=5 to have 5 clients in training (at least 5);
  • the run_client.sh script currently have only GPU-1 in work.

Resources

• Dataset: dataset.fedai.org
• Paper: "Real-World Image Datasets for Federated Learning"

Street_Dataset

We implemented two mainstream object detection algorithms (YOLOv3 and Faster R-CNN). Code for YOLOv3 is borrowed from PyTorch-YOLOv3 and Faster R-CNN from simple-faster-rcnn-pytorch.

• Overview: Image Dataset
• Details: 7 different classes, 956 images with pixels of 704 by 576, 5 or 20 devices
• Task: Object detection for federated learning
• Dataset_description.md

Citation

If you use this code or dataset for your research, please kindly cite our paper:

@article{luo2019real,
  title={Real-World Image Datasets for Federated Learning},
  author={Luo, Jiahuan and Wu, Xueyang and Luo, Yun and Huang, Anbu and Huang, Yunfeng and Liu, Yang and Yang, Qiang},
  journal={arXiv preprint arXiv:1910.11089},
  year={2019}
}