DrawMeA

Inspiration

Everyone wants to draw something at sometime, but what do you do if you can't see the canvas? It's not about being da Vinci, but about showing the emotion and sharing it with people you love (especially in Christmas).

What it does

DrawMeA takes the speech commands as input, and draws objects accordingly on a canvas. The full process is voice-guided.

How I built it

We build on top of Azure Cognitive Service's Speech (LUIS) AI service, and use Python tuned to take specific commands (objects to draw, location to put objects). A pre-defined database contains 10 categories as optional objects to be drawn on the canvas.

Challenges I ran into

The Azure service is not so easy to get in it as a beginner of cloud user. And testing with speech recognition is not Intuitive as classical program debugging.

Accomplishments that I'm proud of

The test run is quiet successful. And all the team members can get familiar with Azure AI services, which we have no clue about how to use. Most importantly, the team works so smoothly and efficiently.

What I learned

Azure AI service framework; great team work needs efforts from everyone.

What's next for DrawMeA

Bring the desktop version to a web platform, which can be easily share to visit by anyone and anywhere; add more categories of objects; add more sophisticated strokes to draw fine details.