Abstract: We propose an efficient abnormal event detection model based on a lightweight masked auto-encoder (AE) applied at the video frame level. The novelty of the proposed model is threefold. First ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results