Problem - Voice chatting online can be very intimidating. Many people are subjected to hateful and offensive language while playing their favorite video games or using various online services.
Solution - For this hackathon we created a program that can recognize voices and censor bad words/hate speech from the person's speech.
Implementation - The person's voice is compared to a list of words/slurs/hate speech and if a match is found the original audio will be replaced by a random positive message spoken by a twitch speech bot.
Scaling for the future - In future implementations, we will use official hate speech and offensive language data from various public social media API’s and use AI to construct more closely related speech alternatives. With increased data quantities, we would implement hash tables to increase our program’s speed and efficiency
https://docs.google.com/presentation/d/1DpP59w48IYxSzL513P2kHoGAKGLnfkBytjaVtP3uDgU/edit?usp=sharing
Log in or sign up for Devpost to join the conversation.