Towards Unified Deep Learning Model for NSFW Image and Video Captioning

The deep learning model is an evolution of an artificial intelligence model called the Artificial Neural Network. And the internal layer of an artificial neural network consisting of layers is a multi-stage structure, the latest deep learning model has a larger number of internal layers, which can result in up to billions of nodes. In addition, learning models combining CNN and RNN to comment on pictures or video are currently being studied. The images or videos are input by CNN, summarized, and the results are input into RNN for printing out meaningful sentences, the so-called image and video captioning. This paper proposes unified deep learning model for NSFW image and video captioning. As noted above, traditional studies on image and video captioning have been approached via a combination of CNN and RNN models. In contrast, in this paper, the classification for safety judgement, object detection, and captioning can all be handled through one dataset definition.