Why train normal file has 10 samples more than original UCF Crime split?

seominseok0429 / Real-world-Anomaly-Detection-in-Surveillance-Videos-pytorch

Real-world Anomaly Detection in Surveillance Videos- pytorch Re-implementation

102 stars 24 forks source link

Open daviduarte opened 1 year ago

daviduarte commented 1 year ago

All the dataset division files is ok, but train_normal.txt has 10 more samples.