[JSAC 2022] VaBUS: Edge-Cloud Real-time Video Analytics via Background Understanding and Subtraction

VaBUS is an edge-cloud video analytics system by leveraging background understanding and subtraction. This repo is the implementation of VaBUS, JSAC 2022.

@ARTICLE{wang2022vabus,
    author={Wang, Hanling and Li, Qing and Sun, Heyang and Chen, Zuozhou and Hao, Yingqian and Peng, Junkun and Yuan, Zhenhui and Fu, Junsheng and Jiang, Yong},
    journal={IEEE Journal on Selected Areas in Communications},
    title={VaBUS: Edge-Cloud Real-time Video Analytics via Background Understanding and Subtraction},
    year={2022},
    volume={},
    number={},
    pages={1-1},
    doi={10.1109/JSAC.2022.3221995}}

---

Environment

Cloud: Python3.8.10 and Ubuntu20.04

Edge: Python3.6.9, C++, and Ubuntu18.04 (Jetson Xavier NX)

Requirement

1. For Python, see src/requirements.txt.
2. For Jetson RoI encoding, you need to rebuild ffmpeg. Please refer to jetson_roi_encode/

Modules and file structures

• dataset/: put your dataset (video or image folder) here
• dataset/ground_truth/: ground_truth results generated by src/generate_encode_size.py and src/generate_inference_res.py for evaluation purpose
• jetson_roi_encode/: for RoI encoding on Jetson
• results/: evaluation results (auto-generated)
• src/cloud.py: main script for cloud running
• src/edge.py: main script for edge running
• src/param.yml: parameter files used by src/cloud.yml and src/edge.yml
• others should be self-explainable

Get started

1. On the cloud server

run python cloud.py

2. On the edge device

run python edge.py

Notice

• Build the ffmpeg library for video encoding on Jetson
• Carefully set up the model files under src/models to adapt to your machine
• We cleaned up the code for clarity.
• Any question please ask. Thanks!