Content monitoring and validation for UGC Platform automated with machine learning

Goal

To automate the process of video validation with minimum moderator involvement.

Obstacle

Video recognition accuracy shall be high.

Solution

We created an ML-powered module based on RCNN (Recurrent Convolutional Neural Network) that can detect inappropriate video content and delete it having notified a user and an admin. The module can detect both video and sound for more accurate content filtering.

Got interested?