Inspiration
Seeing our young relatives and friends, getting mentally and psychologically affected and harmed, from harmful online videos; Violence, Suicide and Harassment. Got us insanely furious, to solve that problem as soon as possible, to save the next era of children.
What it does
TikCare is a web-based mobile application that acts as a real-time, proactive moderation layer for short-video platforms; TikTok. It utilizes an intelligent AI framework to continuously analyze the video feed to detect any harmful content and immediately remove it before ever reaching the user’s screen. Furthermore, it ensures that as the user scrolls, all harmful videos are removed and only safe videos appear. When a threat is detected, we apply some simple changes in the DOM to execute a multi-step, rapid safeguard sequence.
- Visual Obfuscation: Instantly applies an 80px blur to the entire video frame.
- Audio Muting: Silences the video track.
- Interruption & Removal: Pauses the playback, triggers an auto-skip action, and permanently removes the flagged content from the user interface.
IF analyzeFrame(video) returns harmful = TRUE AND confidence ≥ 0.85
THEN applyBlur(80px)
→ muteVideo(true)
→ pauseVideo()
→ autoSkip()
→ removeFromFeed()
How we built it
[HASSONA]
Challenges I ran into
[Later]
