AWS DeepLens is the world’s first deep learning enabled video camera designed to help you grow your machine learning skills by gaining hands-on experience with a physical device capable of running real time computer vision models.

This custom-built device helps you get hands on with deep learning through a series of examples, tutorials, and pre-built models. AWS DeepLens also integrates with Amazon SageMaker, runs AWS Greengrass, and AWS Lambda.

Get creative, and show us what you can build with AWS DeepLens. Choose to customize one of the built-in computer vision models provided, or build and train new models and publish them to the fully-programmable device. You can extend functionality through Lambda functions, calling other AWS services like Amazon Rekognition Video for advanced analysis. The choice is yours. The possibilities are endless.

In this hackathon, put your skills to the test and learn new ones in the process!

AWS DeepLens is putting deep learning in your hands. Machine learning skills within your reach.

What will you build? 

 

Participant Prize: Custom skin to make your AWS DeepLens unique by using your photos, images, and text (valued at $34.95). One per maker who entered an eligible submission

Eligibility

The Hackathon IS open to: 

  • Individuals who are at least the age of majority where they reside as of the time of entry (“Eligible Individuals”)
  • Teams of Eligible Individuals (“Teams”)
  • Organizations (including corporations, not-for-profit corporations and other nonprofit organizations, limited liability companies, partnerships, and other legal entities) that exist and have been organized or incorporated at the time of entry, and employ no more than 50 people (“Small Organizations”).
  • Organizations (including corporations, not-for-profit corporations and other nonprofit organizations, limited liability companies, partnerships, and other legal entities) that employ more than 50 people (“Large organizations”). Please note, however, that Large organizations will only be eligible to win the Large organization Recognition Award, which carries no monetary value. Large organizations will not be eligible to receive any other prize in connection with this Hackathon.
  • Employees of Amazon or its affiliates, Intel, and Devpost are not eligible to participate.
  • NOTE: The Use of AWS DeepLens hardware device is required to enter the Hackathon. Every team (or individual, if submitting as in individual) must have an AWS DeepLens.

Requirements

Main Requirement: Create a working software application that uses and runs on the AWS DeepLens device. Any cloud integrations must use AWS. 

Use AWS DeepLens’s sample projects to get started or build your own custom model from scratch.

Your repository must host the .json model definition, model parameter file, lambda function, gist log and Readme file. The Readme file should contain model location and access instructions, step by step instructions on how to use the trained model and lambda functions, references to any other applicable documents or arxiv papers your project is based on, and testing instructions needed for testing your model. If needed for your solution, also include any side scripts or lambda functions needed to test. 

Hackathon Sponsors

Prizes

$21,024 in prizes

First Place

• $7,500 USD
• Two tickets to AWS re:Invent 2018
• Special edition custom painted AWS DeepLens OR AWS credits ($249 value)

Second Place

• $5,000 USD
• One ticket to AWS re:Invent 2018
• Special edition custom painted AWS DeepLens OR AWS credits ($249 value)

Third Place

• $2,500 USD
• Special edition custom painted AWS DeepLens OR AWS credits ($249 value)

Devpost Achievements

Submitting to this hackathon could earn you:

How to enter

  1. Register for the AWS DeepLens Challenge.
  2. Create an account on AWS.
  3. Set up your AWS DeepLens following the step by set guide.
  4. Visit the Resources page for links to documentation and resources.
  5. Go build! Create your AWS DeepLens model and shoot a short video that demonstrates your masterpiece in action. Prepare a written summary of your project and what it does.
  6. Provide a way to access your project for judging and testing, including a link to your repo hosting the AWS DeepLens code and all deployment files and testing instructions needed for testing your project. (The Github or BitBucket code repository may be public or private. If the repository is private, share access with testing@devpost.com).
  7. Submit your project on awsdeeplens.devpost.com before Feb 14th, 2017 at 5pm EST and be sure to share the links to access to the repo and the deployment files.

Judges

Matt Wood

Matt Wood
Director Product Management, AWS Deep Learning

Jyothi Nookula

Jyothi Nookula
Senior PMT, AWS DeepLens, AWS Deep Learning

Joel Minnick

Joel Minnick
Head, AWS AI Product Marketing

Mike Miller

Mike Miller
Senior Manager, AWS AI

Stefano Soatto

Stefano Soatto
Director of Applied Science, AWS Deep Learning

Jeff Barr

Jeff Barr
Chief Evangelist, AWS

Judging Criteria

  • Quality of the Idea
    Includes creativity and originality of the Application idea.
  • Implementation of the Idea
    Includes how well the idea was executed by the developer.
  • Extent to which the application leverages DeepLens
    Includes how well the developer leveraged the hardware.

Tell your friends

Hackathon sponsors

Brought to you by:
Sponsored by:

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.