machine-learning-augmented/virtual-reality

Inspiration

Art and a desire to experience the world in a different way. This application helps us see the world as we imagine some of the most famous artists may have seen them.

What it does

Features: Leverages Neural networks to apply artistic styles to virtual and augmented reality technologies by using the GPU compute power afforded to us by Amazon Web Services.

How we built it

Cloud Platforms: amazon-web-services - EC2 - GPU compute optimized p2 instances.

Challenges we ran into

AWS instance limits and costs. YouTube file-encoding issues. Initially acquiring large files used for the neural network algorithms.

Accomplishments that we're proud of

Very cool looking VR videos on YouTube as well as prototypes for live Neural Network enhanced AR on a mobile platform.

What we learned

Always learn your AWS limits before you start prototyping.

What's next for machine-learning-augmented/virtual-reality

Fleshing out better processes to create optimal quality videos and images.

Built With

• amazon-web-services
• art
• cloud
• ios
• javascript
• linux
• machine-learning
• neural-networks
• node.js
• python
• tensorf