Build shape
Assistive Android camera app
Low-vision assistive camera app
An Android accessibility product that combines real-time camera assistance and recognition support for low-vision users.

Product brief
Flowy is an accessibility product designed to help low-vision users see more clearly through the camera while also recognizing useful targets in the scene. The challenge was to make enhancement, recognition, and guidance feel like one assistive experience rather than separate features.
In the implemented scope, CameraX capture, OpenGL and GLSL filter processing, bus-number recognition, finger tracking, and voice guidance were connected inside one Android flow. It is a productized example of assistive camera work.
Scope
Problem
The product had to improve what users saw in real time while also recognizing targets and guiding the user through them.
Implementation
I combined CameraX capture, OpenGL/GLSL filtering, bus-number recognition, finger tracking, and voice guidance into one app experience.
Best fit
This fits accessibility cameras and assistive recognition apps that need real-time visual processing and guidance together.
Technical structure
Build shape
Assistive Android camera app
Primary stack
Android · CameraX · OpenGL · PyTorch
Data & sync
Real-time recognition · finger tracking · voice guidance
Operational surface
Bus-number recognition · camera filters · accessibility UX
Demos
Captured flows from the working prototype or production build.
Camera assist flow
Recognition flow
Finger tracking test
Screens
App


Send the current problem, scope, or a reference link. The next step can start from product structure and release scope.