Algorithmia

Introducing Our AWS AMI and DeepFilter Training Pipeline

Algorithmia and AWS There’s been a lot of talk about style transfer lately. Google’s Supercharging Style Transfer research and Facebook’s video style transfer demo are just two examples.

At Algorithmia, we believe in democratizing access to state of the algorithms. That’s why we’re excited to provide an AWS AMI and pipeline for creating your own style transfer models.

You’re now able to setup a deep learning environment in the cloud, train a custom style transfer model, and then make it available as a scalable REST API. Training your own model takes approximately 26 hours end-to-end using Amazon EC2 P2 instances.

We’re also introducing Deep-Style.io, an interactive demo for getting started with style transfer. We’re providing five filters users can apply to their own photos for free.

While we don’t aim to be a photo filter company, using the AMI pipeline above we’ve created 38 different filters developers can use in their apps and services. These are all available via the Algorithmia API.

Deep-Style.io was made using the Deep Filter microservice, our implementation of the Texture Networks research.

Deep Learning on Algorithmia:

Diego Oppenheimer, founder and CEO of Algorithmia.

More Posts - Website

Follow Me:
TwitterLinkedIn