Skip to content

Latest commit

 

History

History
13 lines (10 loc) · 562 Bytes

README.md

File metadata and controls

13 lines (10 loc) · 562 Bytes
Raw

AFT_Codes

Adversarial and Focused Training of Abnormal Videos for Weakly Supervised Anomaly Detection

Fig1

Training

  1. Change the file paths to the download datasets above in list/ucf-train.list and list/ucf-test.list
  2. Feel free to change the hyperparameters in option.py
  3. Train AT and FT modules respectively by changing the parameters in the dataset file and the module in the train file.
  4. run python main.py

Testing

  1. run python late_fusion.py