Research
Searching for App Screens via Interactive Sketching: PSDoodle
Retrieves UI Screenshots from partial sketches
Reduce the time requirement with similar accuracy compared to the state-of-the-art
Methodology: Machine Learning, Computer Vision, Python, JavaScript, AWS
Doodle2App - Native App Code by Freehand UI Sketching.
Generates UI prototype from Low Fidelity Sketches.
Created a Public Corpus of 11k UI Element Drawings compatible with Google QuickDraw Dataset
REMAUI - REMAUI identifies user interface elements such as images, texts, containers, and lists, via computer vision and optical character recognition (OCR) techniques. REMAUI bridge the gap between the digital conceptual drawings of graphic artists and working user interface code.
Generated compilable UI code from screenshots
Extended REMAUI to understand style and incorporated updated OCR technology.
Methodology: Computer Vision, OCR, Python