abnormaly detection
combinated traditional method to detect abnormaly detection


Figure 1. Labeling and preprocessing for abnormal behavior.


Figure 2. Action (Abnormal) detection with 3D data.


Figure 3. Abnormal detection from Sensedrive code.
-
References
(1) https://arxiv.org/pdf/1801.04264.pdf -> ucf crime (2) https://arxiv.org/pdf/2003.01920.pdf -> ETRI 3D Action (3) https://arxiv.org/pdf/1412.0767.pdf -> C3d (4) https://rose1.ntu.edu.sg/dataset/actionRecognition/ -> NTU RGB-D dataset (5) https://www.mdstech.co.kr/SenseDMS -> hancom mds SenseDMS