Sketch Mentor

Main Page
Canvas AI
Youtube Transcriber
Handwritten AI
Transcription

Inspiration

In rural areas, access to quality educational resources, especially in art and creativity, is limited. We wanted to create a platform that nurtures artistic talent by providing AI-powered mentorship, helping students enhance their drawing skills without needing constant teacher supervision.

What it does

Sketch Mentor is an AI-driven web application that guides students through sketching exercises. Users upload or create sketches, and the system provides real-time feedback, improvement suggestions, and practice exercises based on their skill level. It also recommends learning paths tailored to individual progress.

How we built it

We built Sketch Mentor using:

Frontend : React Web for an interactive and responsive UI

Backend: Python with FastAPI for API development

Machine Learning: A deep learning model trained on sketches to evaluate and suggest improvements

Database: MongoDB for user profiles and progress tracking

Deployment: Render and Vercel for hosting the backend and frontend respectively

Challenges we ran into

Building an ML model that accurately evaluates artistic sketches was harder than expected. Ensuring the system runs smoothly on low-bandwidth networks to support rural education settings. Integrating real-time feedback without causing delays in the user experience.

Accomplishments that we're proud of

Successfully deploying a working prototype that can evaluate and give meaningful feedback on sketches Creating an intuitive, beginner-friendly user interface that even young students can navigate easily. Building an end-to-end solution from scratch within the hackathon timeframe.

What we learned

How to optimize machine learning models for inference speed. How to design an educational application that is effective even with minimal internet access. The importance of user-centric design when working with young or first-time tech users.

What's next for Sketch Mentor

Adding multi-language support, especially for regional languages like Tamil. Expanding feedback types to include shading, perspective, and color techniques. Collaborating with rural schools to pilot Sketch Mentor in real-world classrooms. Improving the ML model with more diverse datasets to handle various art styles.