elsheikh21/venice-boats-classification

using MarDCT dataset to distinguish between different types of venice boats

Venice Boats Classification

Goal: Implement a ML algorithm to differentiate between 24 different classes (23 different types of boats, and 1 class for water)

Libraries Used

this repo runs on python 3.6.4(x64), and pip (18.1). You can install the python requirements with
python -m pip install -r requirements.txt

    git clone https://github.com/elsheikh21/venice-boats-classification
    python -m pip install -r requirements.txt

Dataset

• Downloaded from (here)

• 4774 training images, 1969 testing images

  • the dataset was altered, to get a number of matching classes between the training and testing classes. So we end up with 19 classes.  

  • Note folders removed names are hard coded in the code for that purpose

---

How it works

Check docs/how_it_works.md

---

Output

Check the report

---