Text this: Integrating 3D convolutional neural network and Bidirectional Encoder Representations from Transformers for effective video summarization.