Repository for the CS 452 group working on the Ghost Cat project
Magpie Labs has spun out a service called Ghost Cat Analytics (ghostcat.io), to analyze the content of remote camera images used to track wildlife in large-scale studies. The service outputs predictions of what species and how many of each are in images. Previous clients have mostly just required large .csv summaries of the contents of their images, but future clients will want a web interface.
The goal of the project will be to build a web application that will enable wildlife researchers to store their data. The app could do things such as: Provide a GUI for uploading new data into a database Provide a GUI for fetching images and filtering them by date ranges and species, as defined in the image metadata and ML model outputs.
Currently, a lot of data is stored on Amazon s3 buckets, but we intend our project to store the data using AWS databases. For the NoSQL database, this means we anticipate using DynamoDB.
We have been given permission to use data from Ghost Cat Analytics by David Healy (an employee at Ghost Cat). This information is also not super confidential, because it consists mostly of images of wildlife, so we do not anticipate running into too many issues regarding data privacy.
While this may seem like a niche project, the business value of this product is quite large. Our team member at Ghost Cat has many users that are anxiously awaiting this software integration to allow for easier access to their photos for simplified population tracking.