/huaweicloud_2023

An Implementation of Fatigue Driving Detect, Huawei Cloud Track, 18th Challenge Cup

Primary LanguagePythonApache License 2.0Apache-2.0

18th Challenge Cup National College Students' Extracurricular Academic Science and Technology Works Competition - "Revealing the List and Taking Command" Special Competition · Huawei Cloud Track Second Prize

Competition Homepage

Competition Homepage

Certificate

Certificate

Certificate Website

Project Overview:

The cloud-based ranking algorithm utilizes yolov5 + Openvino + Huawei Cloud ModelArts. It employs binary search and divide-and-conquer algorithms for optimization, allowing for the detection of fatigue states' start and end times without traversing the entire video.

The edge-side (Jetson TX2 NX) algorithm uses Deepstream (C/C++ 6.0.1) + yolov5. Specific implementation codes include:

These implementations are in C/C++, as the Python implementation, according to NVIDIA, is currently unfeasible.

Refer to the technical documentation in this directory for detailed implementation.

Team Name: The Big Radish of the Production Team
Program Leader: Gongbo Zhang
Directors: Jian Zhou, Fei Wu
Teammates: Shuming Guo, Luran Lv, Aolin Zhang, Xingyu Chen, Jintian Wu, Yufan Jia, Zheyu Zhou, Jiahao Zhang, Jinshen Zhang

Acknowledgments: Minhan Tang, Yongye Lai, Haoyu Deng, Shiyu Zhang