Neural network protects women from unwanted penis photos
Seattle programmer Kelsey Bressler announced the start of work on a neural network that will be able to analyze photos sent to personal messages on social networks and delete genital images (dikpics) before the recipient sees them, reports BBC.
The programmer thought about the need to create such an algorithm after she received another dikpik in private messages on Twitter. After studying the reviews of other users of the social network, the girl realized the massiveness of the problem and, together with her colleagues, decided to develop a neural network.
To train the program in Twittere, a collection of intimate photos from everyone was announced in private messages, and now a team of specialists is teaching software to recognize dikpiks and delete them. After developing the neural network, Kelsey Bressler is ready to provide her Twitter algorithm so that they can embed it in the social network.