We propose a live streaming platform that facilitates real-time content moderation in the form of censoring audio and chats. This is powered by our toxicity filtration service which could be embedded into any existing content platform. The service provides filtration based on disparate classification filters.
- Endpoints for detecting and filtering abusive and offensive content a
- Endpoint for predicting the type of sentence from the 6 categories
- Custom filtering words can be added to the endpoints
- Reporting feature for collecting more data for the model
- Integration with Agaro.io for Content moderation on live streams
- Integration with a chat application
- Live visualization of sentences and its class
We're are open to enhancements & bug-fixes 😄
- This project was made at AngelHack 2019, Mumbai under
36 hours with minimal pre-preparation