Moment Scraper

Moment Capture Overview
Capture with a finger frame
Captured moment (Just Photo)
Moment Viewer (Photo and Ambient Sound)

Inspiration

While spending time with my child, there were many moments when I wanted to capture exactly what was happening. Taking out a smartphone was often too slow, but continuously recording everything with a vlog camera also did not feel right. I wanted to create a way to intentionally capture only the moments that matter.

What it does

Moment Scraper is a wearable camera that lets you capture a moment by making a frame with your fingers. When the gesture is detected, it records both a photo and the ambient sound around you. The goal is to preserve not only what you saw, but also what it felt like to be there.

How we built it

We prepared a camera, microphone, speaker, processing unit, and battery. Then we asked Codex to help us build the rest. It supported us through the whole process, from writing and debugging the code to improving the interaction between the hardware and software.

Challenges we ran into

Detecting the finger-framing gesture was more difficult than expected. Because of the limitations of the ESP32 hardware, it was challenging to improve recognition accuracy while also keeping the latency low.

Accomplishments that we're proud of

We built fast, accurate finger-frame detection on an ESP32, and did it with GPT-5.6 and Codex without writing a single line of code ourselves.

What we learned

We learned that having a clear vision and turning it into reality are still valuable human roles. AI can handle much of the execution, but clearly communicating the scope and desired level of polish is essential.

What's next for Moment Scraper

There is still a lot to improve before it becomes practical, including gesture recognition, image quality, and stabilization. We want to rethink the whole technology stack by using camera glasses and moving some processing to a smartphone, and eventually bring this experience to people around the world.