-
-
logo
-
ios app
-
web app
-
web app
Inspiration
According to the National Center on Elder Abuse, nearly 30% of individuals with disabilities who depend on daily care experience some form of mistreatment. Additionally, the National Council on Aging estimates that 1 in 10 Americans over the age of 60 experiences elder abuse. Since many such cases go unreported, the real statistics are higher. Currently, no systems exist that can monitor the live feed for you, requiring guardians of vulnerable individuals to manually monitor the footage.
What it does
We have built EyeQ, which is an intelligent system that helps flag abuse by a professional caregiver. This can be used to notify guardians of individuals who are unable to communicate, such as those with cognitive disabilities like cerebral palsy, the elderly with conditions like dementia, and children. The aim is to prevent such occurrences through vigilance, but the live feed is stored and can be used to indict the caregiver, too.
How we built it
We used YOLO-11 for pose estimation. If YOLO flags suspicious movement, the video is further sent to Gemini API for classification. If Gemini detects any sort of physical violence, the application sends a push notification to the mobile app, built using SwiftUI.
Challenges we ran into
The call to Gemini and the outcome are sensitive to the prompt used. We had to tune our prompt to guarantee reliable behavior from the application.
Accomplishments that we're proud of
This is the first-ever intelligent system for caregiver abuse detection. It makes the task of watching over your loved ones easier than ever before.
What we learned
We learned to parallelize our workflow among members and integrate them all into one ecosystem.
What's next for EyeQ
We aim to incorporate microphones and mechanisms for detecting verbal abuse into our system.
Log in or sign up for Devpost to join the conversation.