One More Description

Main webpage of One More Description
Step1: Creat a Character.
Step2: Creat a room
Take a guess of the description
Now it is your trun to choose the word
Now it is your turn to describe the word
Final Score
Scheme

Inspiration

Inspired by projects like QuickDraw where valuable data is collected while people are having fun, we decided to create a similar experience that is beneficial for scientific research and technology development.

What it does

“One more description” is an online web game, where the players explain a word to each other using terms aside from that word by typing or speaking out loud. The game collects those descriptions and builds up a dataset, which is beneficial for:

Semantic research
For instance, semantic research about the perception of words across different cultures and backgrounds, so this kind of dataset could help gain more insights about language development and word meanings.
Virtual assistants and search engines Have you been in a situation where you wanted to say an artist, company name, or other word but you couldn't remember it? It's a global phenomenon called the "Tip of your tongue" feeling. Imagine that you could describe a word sitting on tip of your tongue to a virtual assistant or other search engines and they will tell you that word. By having a multiple descriptions dataset we can help to achieve that!

How we built it

We created the mockup in Figma and tried the Web-Speech-API on the local server.

Challenges we ran into

We have only basic knowledge about machine learning, so we tried to find a source of validation if the definition data is really valuable for people.

Accomplishments that we're proud of

We are proud of our clear task distribution during the teamwork, with which we worked efficiently and effectively, the design style, and the idea.

What we learned

We learned more about Machine Teaching, we explored the possibilities of voice recognition and played with different javascript libraries that can do that, to check their reliability. We were really surprised by how well speech recognition works nowadays!