Official PyTorch implementation of "Synthetic Temporal Anomaly Guided End-to-End Video Anomaly Detection"

This is the implementation of the paper "Synthetic Temporal Anomaly Guided End-to-End Video Anomaly Detection" (ICCV Workshops 2021: RSL-CV).

Paper || Presentation Video

Dependencies

• Python 3.6
• PyTorch = 1.7.0
• Numpy
• Sklearn

Datasets

• USCD Ped2 [dataset]
• CUHK Avenue [dataset]
• ShanghaiTech [dataset]

Download the datasets into dataset folder, like ./dataset/ped2/, ./dataset/avenue/, ./dataset/shanghai/

Training

git clone https://github.com/aseuteurideu/STEAL

• Training baseline

python train.py --dataset_type ped2

• Training STEAL Net

python train.py --dataset_type ped2 --pseudo_anomaly_jump 0.2 --jump 2 3 4 5

Select --dataset_type from ped2, avenue, or shanghai.

For more details, check train.py

Pre-trained model

Evaluation

• Test the model

python evaluate.py --dataset_type ped2 --model_dir path_to_weight_file.pth

• Test the model and save result image

python evaluate.py --dataset_type ped2 --model_dir path_to_weight_file.pth --img_dir folder_path_to_save_image_results

• Test the model and generate demonstration video frames

python evaluate.py --dataset_type ped2 --model_dir path_to_weight_file.pth --vid_dir folder_path_to_save_video_results

Then compile the frames into video. For example, to compile the first video in ubuntu:

ffmpeg -framerate 10 -i frame_00_%04d.png -c:v libx264 -profile:v high -crf 20 -pix_fmt yuv420p video_00.mp4

Bibtex

@InProceedings{Astrid_2021_ICCV,
    author    = {Astrid, Marcella and Zaheer, Muhammad Zaigham and Lee, Seung-Ik},
    title     = {Synthetic Temporal Anomaly Guided End-to-End Video Anomaly Detection},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops},
    month     = {October},
    year      = {2021},
    pages     = {207-214}
}

Acknowledgement

The code is built on top of code provided by Park et al. [ github ] and Gong et al. [ github ]