The unicorn will always find who did it - as long as you provide the data!

• Step 1:

  Clone the repo.

• Step 2:

  In a terminal go to backend directory and run pip install -r requirements.txt (windows) or pip3 install -r requirements.txt (mac).

• Step 3:

  In the backend directory, run python result.py (windows) or python3 result.py (mac). The backend server should start at localhost:8080

• Step 4:

  Change the directory to unicornlab and run npm install in the terminal

• Step 5:

  In the unicornlab directory, run npm run dev. The frontend server should start at localhost:3000

• Step 6:

  In a browser, open localhost:3000 and start the investigation!

The uploaded .bmp files first go through pre-processing to increase ridge detection and decrease noise. Then we use Lowe's ratio test and RANSAC homography to calculate the simillarity scores and normalize them.

The upload .fatsa files are parsed, and then we use the Smith-Waterman local alignment method (N bases as wildcards) to calculate an accurate score and then normalize them.