National Day of Civic Hacking

_{Following description comes from [Code for America](https://www.codeforamerica.org site.}

Code for America, a “network of people making government work for the people, by the people, in the 21st century”, annually encourages interested people to participate in a National Day of Civic Hacking project.

The National Day of Civic Hacking is a nationwide day of action to bring together civic leaders to work with our local governments to tackle some of our toughest challenges.

Some types of successful events from the past have been:

• Hackathons: Events where people come in with ideas for an app and create prototypes at the event.
• City Camps: Unconference [sic] events where people discuss the problems in their city and create the connections necessary to solve them
• Design Camps: Events where we use user-centered design to reimagine government services.

US FWS and USGS Patuxent 2017 Project

In August 2017 a hackathon event was held at the PWRC with the goal of redesigning and modernizing the Harvest Survey website.

(Description to be added)

Comic Inspiration

_{more info about comic}

Teach the Machine

Our primary goal is to develop a workflow for photo tagging to support aerial digital survey efforts and support AI/ML research efforts. Focusing on an app/interface for humans to generate training sets of images for the AI/ML process.

AI/ML techniques require teaching the machine what we see. It is truly an education by example. We feed it lots of images and indicate which one contains targets (birds in this case) and which ones don’t. Through the magic of math and computer stuff, the AI learns which pictures contain targets and which don’t. Our goal is to construct a method for human viewers to annotate images. The initial development will focus on tagging photos that do and do not contain birds. The resulting photo sets can be used to create an algorithm to automatically identify images with targets (birds).

Our goal is to build a method to display sections of large images to a user and allow the user to indicate whether (and possibly where) they see birds; these results then need to be summarized to classify the original image. Most of our user and programmers are familiar with R, so it may be a natural language for parts of this project, while Python would also be suitable for some aspects.

The Machine Uprising

The forward-looking stretch goal is to set up a functioning process to filter images based on the training set created from the first goal.

We want to be lazy and have the machines do our work for us. We are hoping to have access to some AI/ML resources to play with existing technologies. Possibilities include … (to be filled in as we are successful at borrowing/begging).