Machine Learning Project (Bike classification)

All images taken from 99Spokes.com.

Goal of this project is to classify pictures of bikes using a neural network. The AI should be able to decide whether a bike is a full suspension bike or a hardtail. A secondary goal is to correctly classify e-bikes and bio bikes.

How to download new images (requires Node.js)

Images have been added to the source code. There is no need to run this step other that to download newly added images.

Requirements

• Node v17.5 or higher (fetch api required)

Getting images

Images are being fetched from 99spokes.com as they provide lots of images of bikes with white background.

Download Images

make sure that the page url, selector and folder paths are configured propperly.

Make sure you are running at least version 17.5 of Node.js Then install the packages with the following command:

npm install

To get image urls the following command: They should now be in the folder ./urls/

node 01_GetImageUrls.js

To download images from those urls run: They should now be in the folder ./images/

node 02_DownloadImages.js

Urls and images have been organized in folders according to their labels.

Running the Project file

To run the project file you need to have python3 and jupyter notebook installed. Open 03_Project.ipynb and click the "Run All"-Button. This step is not required because the previous outputs have been saved.

---

Links provided by teacher

• https://tensorflow.rstudio.com/tutorials/beginners/basic-ml/tutorial_basic_classification/

• https://www.r-bloggers.com/2021/03/how-to-build-your-own-image-recognition-app-with-r-part-1/

• https://www.r-bloggers.com/2020/10/platypus-r-package-for-object-detection-and-image-segmentation/