Inspiration
Nearly 3 million US children experience some form of abuse. Around 25 million images are reviewed by the national center for missing & exploited children. We wanted to build technology to automate this process.
What it does
we built an advanced video recognition system where a user inputs the URL of the webpage he thinks has child pornography. The script fetches the videos from the page and performs analysis of emotion, age, and obscenity of the video frames to predict if the video has any child victim
How we built it
We Used Azure cloud services and Microsoft vision API's services to detect age, gender, emotion, and other facial features. We built a platform to generate data for future researchers in this domain.
Challenges we ran into
To deploy our scripts on the cloud and integrate several api's together
Accomplishments that we're proud of
Inspite of not having any experience with Azure we managed to build a decent model to detect child abuse.
What we learned
AWS , azure , open CV, WIX
What's next for Netizens Against Child Abuse
To improve ML models and improve the platform
Log in or sign up for Devpost to join the conversation.