You can see in Frame by Frame file we can implement a function which tells if video is safe for kids or not and We can take actions accordingly.

Mobile Net Architecture is used for the classification as it is very fast and has less params to train.

This is inspired by https://www.freecodecamp.org/news/how-to-set-up-nsfw-content-detection-with-machine-learning-229a9725829c/ , https://github.com/GantMan/nsfw_model The Data for this project is collected with help of scripts at https://github.com/alex000kim/nsfw_data_scraper