ClarityCam

• 

  Screenshot of ClarityCam

Inspiration

I've always been a messy person. If I were to describe myself, it would be having the "Sidam Touch" (a play on words for Midas Touch), where everywhere I go gets cluttered with random objects. From days-old cups, tissues, and papers littering the floor, it's an understatement to say I'm "disorganized". Running around the house looking for my wallet, when it's been sitting on my desk (despite searching said desk 5 separate times), has been a common occurrence to my parents. Although I can't say I'm tidy, I make an effort to fix this bad habit of mine (in hopes that I save time getting out the door, so I don't keep end up being late)

What it does

ClarityCam is a camera-based mobile app that helps users identify and isolate objects in messy or cluttered scenes. As of right now, it allows users to take a photo, quite similar to a camera app! It stores the image in their gallery/photo app, which will eventually be processed through an object detection model. The goal is to let users select a specific object from a list and have the app highlight every instance of that object in the image—helping them find what they need faster! ** Right now, it's only a camera app-

How we built it

I started by building a basic mobile app with camera functionality. Users can open the app, preview the live feed, and take a photo with a simple button press. The UI is structured to support an object detection pipeline in the next phase, with plans to incorporate image segmentation models like YOLOv8 for real-time visual parsing. The frontend is built using Flutter for cross-platform support, which was tested on a physical device (my Android phone, as a test subject).

Challenges we ran into

I did not actually implement the object detection part of the program due to lack of time. The main challenges were downloading and installing and figuring out how Flutter, Android Studio, Yolov5 (the object detection model I was planning on implementing), worked.

Accomplishments that we're proud of

My mobile app, on a physical device. It can ACTUALLY take pictures. I made something useful ...

What we learned

The world is big (and has lots of errors). Flutter cool, Yolov5 cool (but I need better time management skills), ChatGPT can be cool ... when it doesn't give me more errors ...

What's next for ClarityCam

First, I would like to add the Object Detection part onto the program! Next, I really want to make it more similar to the camera app, with pressing the button causing this sort of vibrate noise. I want to experiment on other phones (such as iOS)! Lastly, I want to include image segmentation masks, where taking a picture automatically applies the object detection and allows the user to scroll through which masks they want to see. To view the different objects, I also want to use the carousel feature for each object!

Built With

• dart
• flutter