Sketch Mentor

Make Education Easier

Comment

🎨 Inspiration

Access to quality art education is still limited in many rural and underserved areas. Many talented students lack mentors who can guide their artistic growth. Sketch Mentor was born from the idea of bridging that gap — bringing personalized, high-quality art feedback and learning opportunities to every learner, regardless of location or resources.

💡 What it does

Sketch Mentor is an AI-powered art education platform that acts like a virtual art tutor.
Students can draw digitally or upload hand-drawn sketches, and the system instantly provides personalized feedback on:

  • Accuracy and proportions
  • Structure and composition
  • Artistic expression

It also:

  • Tracks progress and improvements over time
  • Adapts to each student’s learning path
  • Works efficiently even in low-bandwidth or offline environments

🛠️ How we built it

Frontend:

  • Built using Flutter Web (Dart) for a smooth, responsive, and cross-platform UI.

Backend:

  • Developed with FastAPI (Python) and hosted on Render for reliability and scalability.

Machine Learning & Analysis:

  • Used TensorFlow/PyTorch and OpenCV for image evaluation (shape detection, proportion analysis, line quality, etc.).
  • Models trained and tested in Google Colab.

Database & Hosting:

  • MongoDB for storing user data, feedback history, and progress metrics.
  • Frontend hosted on Vercel.

Design & Collaboration Tools:

  • UI/UX mockups in Figma and Canva
  • Version control via GitHub
  • Testing and API management through Postman

🚧 Challenges we ran into

  • Handling diverse sketching styles and varying drawing quality.
  • Building a responsive ML model that provides real-time feedback.
  • Ensuring offline functionality for rural environments.
  • Designing an intuitive UI for students with limited digital familiarity.
  • Managing efficient data synchronization under low connectivity conditions.

🏆 Accomplishments that we're proud of

  • A fully functional prototype with end-to-end integration.
  • Real-time sketch assessment that simulates a mentor-style feedback experience.
  • A design optimized for low-bandwidth, rural connectivity.
  • Positive initial testing results showing improved engagement and learning confidence.

📚 What we learned

  • The importance of accessibility-first design in educational technology.
  • How to fine-tune ML models for subjective visual evaluation tasks like art.
  • The value of cross-functional collaboration across design, AI, and UX.
  • How AI can empower—not replace—human creativity and mentorship.

🔭 What's next for Sketch Mentor

  • Adding regional language support (starting with Tamil).
  • Expanding AI feedback to include shading, perspective, and color analysis.
  • Introducing gamified learning — badges, challenges, and progress milestones.
  • Partnering with NGOs, rural schools, and government programs to pilot large-scale deployment.
  • Building a community-driven art learning ecosystem to nurture creative growth worldwide.

Built With

Share this project:

Updates