Table 9 Complexity comparison of models.
From: Multimodal anomaly detection in complex environments using video and audio fusion
Case number | Scene description | Real label | Model prediction | Prediction result | Cause of failure |
|---|---|---|---|---|---|
Case 1 | The flash brightness of vehicles changes dramatically. | Abnormal | Normal | False negative | Under the rapidly changing lighting conditions, the model failed to capture the abnormal behavior. |
Case 2 | Angle similarity | Abnormal | Normal | False negative | The similarity of objects’ behaviors makes the model unable to detect potential anomalies. |
Case 3 | Complex background | Normal | Abnormal | False positive | The background forest causes too much noise, and the model misjudges normal activities as abnormal behaviors. |
