Unbox AI

welcome to subscribe my channel

• youtube channel
• bilibili channel

unbox project

• Zero-VIRUS*: Zero-shot VehIcle Route Understanding System for Intelligent Transportation (CVPR 2020 AI City Challenge Track 1)

video

• youtube

system requirements

• ubuntu 18.04
• python >= 3.6
• cuda 10.2

setup environments

1. clone source codes

   $ git clone https://github.com/dyh/unbox_Lijun-Yu_zero_virus.git

2. enter project directory

   $ cd unbox_Lijun-Yu_zero_virus

3. create a python virtual environment

   $ python3 -m venv venv

4. activate the virtual environment

   $ source venv/bin/activate

5. upgrade pip

   $ python -m pip install --upgrade pip

6. install requirements package

   1. install other packages

      $ pip install -r requirements.txt

   2. install detectron2 (based on cuda 10.2 and pytorch 1.5.1)

      and you could choose other version at here

      $ python -m pip install https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.5/detectron2-0.2.1%2Bcu102-cp36-cp36m-linux_x86_64.whl

execute the program

1. detect the "./unbox_test/input/MVI_40855.mp4" video file of this project

   $ python unbox.py

2. the output results are saved in "./unbox_test/output" directory for some image files

   0.png, 1.png, 2.png ... n.png

3. you can use ffmpeg to merge these images into one video file

   $ ffmpeg -f image2 -i ./unbox_test/output/%d.png ./unbox_test/output.mp4

sample dataset

• the test video MVI_40855.mp4 is made up of images in the directory "MVI_40855" in the DETRAC-test-data.zip file.
• download dataset: http://detrac-db.rit.albany.edu/download
• introduction of dataset: http://smart-city-sjsu.net/AICityChallenge/data.html

---

AI开箱

欢迎订阅我的频道

• bilibili频道
• youtube频道

开箱项目

• 用于智能交通的 Zero-shot 车辆路线理解系统
• Zero-VIRUS*: Zero-shot VehIcle Route Understanding System for Intelligent Transportation (CVPR 2020 AI City Challenge Track 1)

视频

• bilibili

系统需求

• ubuntu 18.04
• python >= 3.6
• cuda 10.2

环境配置

1. 下载代码

   $ git clone https://github.com/dyh/unbox_Lijun-Yu_zero_virus.git

2. 进入目录

   $ cd unbox_Lijun-Yu_zero_virus

3. 创建python虚拟环境

   $ python3 -m venv venv

4. 激活虚拟环境

   $ source venv/bin/activate

5. 升级pip

   $ python -m pip install --upgrade pip

6. 安装软件包

   1. 安装其他包

      $ pip install -r requirements.txt

   2. 安装 detectron2 (基于 cuda 10.2 和 pytorch 1.5.1)

      你也可以在 这里 选择其他版本

      $ python -m pip install https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.5/detectron2-0.2.1%2Bcu102-cp36-cp36m-linux_x86_64.whl

运行程序

1. 对项目中的 ./unbox_test/input/MVI_40855.mp4 视频文件进行检测

   $ python unbox.py

2. 输出结果为图片文件，保存在 ./unbox_test/output 目录

   0.png, 1.png, 2.png ... n.png

3. 可以使用 ffmpeg 将图片文件合并为视频文件

   $ ffmpeg -f image2 -i ./unbox_test/output/%d.png ./unbox_test/output.mp4

样本数据集

• 测试视频 MVI_40855.mp4 是由 DETRAC-test-data.zip 文件中的 MVI_40855 目录下图片组成。
• 数据集下载 http://detrac-db.rit.albany.edu/download
• 数据集说明 http://smart-city-sjsu.net/AICityChallenge/data.html

---

below is the origin README file

---

Zero-VIRUS*: Zero-shot VehIcle Route Understanding System for Intelligent Transportation (CVPR 2020 AI City Challenge Track 1)

Authors: Lijun Yu, Qianyu Feng, Yijun Qian, Wenhe Liu, Alexander G. Hauptmann
Email: lijun@lj-y.com

*Written in the era of Coronavirus Disease 2019 (COVID-19), with a sincere hope for a better world.

@inproceedings{yu2020zero,
  title={Zero-VIRUS: Zero-shot VehIcle Route Understanding System for Intelligent Transportation},
  author={Yu, Lijun and Feng, Qianyu and Qian, Yijun and Liu, Wenhe and Hauptmann, Alexander G.},
  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops},
  year={2020}
}

Setup

Install miniconda, then create the environment and activate it via

conda env create -f environment.yml
conda activate zero_virus

Directory structure:

• datasets
  • Dataset_A (AIC20_track1_vehicle_counting.zip/Dataset_A)
  • Dataset_B (hidden evaluation)
• experiments
  • efficiency
    • aic2020-base.json
  • <experiment_name>
    • output.txt

Evaluate

As a zero-shot system, no training is required. We use Mask R-CNN pretrained on COCO from detectron2 as detector, whose weights will be downloaded automatically at the first run.

As the dataset only provided screenshots of the pre-defined routes, we created our own annotation of them with labelme.

To get system outputs, run

./evaluate.sh <experiment_name> <dataset_split>
# For example
./evaluate.sh submission Dataset_A

To get efficiency base score, run

python utils/efficiency_base.py

Performance

On Dataset A with 8 V100 GPUs:

• S1: 0.9328
  • S1_Effectiveness: 0.9120
    • mwRMSE: 4.2738
  • S1_Efficiency: 0.9815
    • time: 3084.04
    • baseline: 0.546801

Visualizations available at Google Drive.

License

See LICENSE. Please read before use.