Chem-o-Crypt

Inspiration

This is my PhD dissertation project.

What it does

Chem-o-crypt is a platformer game designed to teach chemical equation balancing or encryption/decryption using Caesar cipher. During each level, player has to run around the platform to collect correct number of chemical elements required to balance the equation. In the cryptography version, instead of chemical elements, player is tasked with collecting the letters of the cipher/plain text corresponding to the desired encryption/decryption. Game mechanic involves patrolling enemies, flying bats, jumping, moving around, moving platforms and some hazards such as water and thorns. Player needs to collect items while avoiding enemies and hazards. The game adapts dynamically to the skill and emotions of the player predicted using a Dynamic Bayesian Network and facial emotional analysis, respectively.

How I built it

I built it using Unity, Bayesserver and Affectiva (from iMotions)

Challenges I ran into

Used the basic facial expression to predict the emotions of frustration, boredom and engagement during a video game play. Accurately getting the facial expression was a challenging task, as it involves making the people sit in front of their computer for long time in proper lighting conditions before their patience runs out. Learning the parameters of the Bayesian Network was a challenging task as well, because it needs a large amount of data to learn the conditional probabilities of the nodes in the network.

Accomplishments that I'm proud of

I was able to do everything on my own without major help from anyone. I did receive help from my advisors regarding the theoretical part, but I implemented it all on my own.

What I learned

I learned a lot about human facial expression as well Dynamic Bayesian Networks.

What's next for Chem-o-Crypt

The game is being utilized for further experiments that can establish the effectiveness of a content agnostic game based architecture also called as CAGE (https://repository.asu.edu/attachments/186498/content/Baron_asu_0010E_16979.pdf).

Built With

• affect
• c#
• unity