Hybrid deep learning model for suspicious human-action recognition.

In this paper, a video classifier using transfer learning with VGG16 (Convolutional neural network model pre-trained on the ImageNet-1k dataset) base model and hybrid transformers on the UCF Crime Dataset of 100GB is presented. Thisdeep learning approach is presenting a hybrid model for video classification having applications in scene identification, video classification, security, and many. [ABSTRACT FROM AUTHOR]