Welcome to MemoryCare 🧠💫

A Gentle AI Companion to Help Prosopagnosia Patients Recognize Their Loved Ones

The Challenge 💔

When Memory Fades, Connection Breaks: For people in the early stages of memory loss or facial recognition issues, those moments happen more often than they’d like to admit. Whether it’s forgetting a co-worker’s name or hesitating to greet a neighbor you’ve known for years, it’s frustrating—and sometimes isolating. Prosopagnosia (face blindness) and early cognitive decline don’t always mean total memory loss. Often, it’s a matter of needing a little help reconnecting the dots in day-to-day life.

The Heart of Our Solution ❤️‍🩹

Enter MemoryCare - A compassionate AI assistant that bridges the gap between memory and reality. Using real-time face recognition and a personalized memory database, it gently reminds patients who’s in front of them—even multiple people at once—, whether it’s a daughter, a best friend, or a caregiver. It doesn’t stop at names: it recalls moments, shared experiences, and emotional context—restoring identity and dignity, one recognition at a time.

How MemoryCare Works 🧠🎥

Memory-Care runs live on a device with a camera (like a laptop or smart screen). Here’s what happens step-by-step:

  1. Live Face Detection
    When the app is running, it uses the camera to detect faces in real time. A box appears around each detected face—just like in a video call.

  2. Face Recognition & Labeling

    • If the face is not recognized, it will simply label the box as "Unknown Person."
    • If the face matches someone already stored in the user’s People Database, the label beside the box updates to show the person’s name (e.g., “Sarah”).
  3. Context-Aware Audio Prompt
    As soon as a match is confirmed, the system plays a short audio message to help the user recognize who they’re seeing.

    “This is Sarah, your daughter. You had lunch together last Sunday.”
    The message is generated using personalized facts and memories stored for that person.

Key Features 🌟

  • Real-Time Face Recognition: Works with multiple people in a single frame.
  • Personal Memory Recall: Brings back shared memories, not just names.
  • Voice-Optional Interaction: Silent or voice-activated, depending on your preferences.
  • Discreet, Respectful Design: Never intrusive or overwhelming.
  • Privacy-First: All data is securely stored and accessible only to you.

Built With 🛠

  • Next.js: Fast, responsive UI with server-side rendering.
  • Tailwind CSS: Customizable, sleek styling for a user-friendly interface.
  • Supabase – User authentication, face image storage, memory database
  • FaceAPI – Multi-person detection and embedding
  • Gemini Live – Natural memory summaries and audio responses

Who It’s For 👥

Familiar Faces is designed for people who’ve started to notice memory shifts and want to stay ahead:

  • “I used to be great with faces. Lately, not so much.”
  • “I don’t want to keep asking who someone is.”
  • “I’d rather have a private tool than rely on others all the time.”

Whether you're navigating early-onset Alzheimer’s, age-related memory fog, or prosopagnosia, this gives you more control over your connections.

What’s Next for MemoryCare✨

We’re exploring ways to bring MemoryCare into daily life through a wearable companion, such as smart glasses or a discreet clip-on camera, allowing real-time recognition and memory prompts on the go. Also building a memory timeline feature that lets users review who they interacted with each day—helping reinforce routine, jog memory, and reflect on meaningful moments.

Why MemoryCare?

Because memory isn’t just about data—it’s about the connection we shared and cherished.
MemoryCare doesn’t replace your memory; it helps support it. It’s your quiet companion in the background, helping you feel more present, more confident, and more you.

You’re not broken. You’re human.
And this is here to help you feel remembered—even when your memory needs a little help.

Built With

+ 6 more
Share this project:

Updates