C2W4_Lab1 Rock Paper Scissors dataset is not accessible

It appears Laurence Moroney’s website is down, so the ‘Rock Paper Scissors’ dataset is not accessible at the links provided. The notebook is accessible via the GitHub repo, however there is no dataset available right now (18 July 2023).

http://www.laurencemoroney.com/rock-paper-scissors-dataset/

https://storage.googleapis.com/laurencemoroney-blog.appspot.com/rps.zip

https://storage.googleapis.com/laurencemoroney-blog.appspot.com/rps-test-set.zip

https://storage.googleapis.com/laurencemoroney-blog.appspot.com/rps-validation.zip

The links given in the post don’t match the ones mentioned in the notebook.

For instance, the train set is linked via:
!wget https://storage.googleapis.com/tensorflow-1-public/course2/week4/rps.zip

Please pull the latest changes in the repo to your local copy and try again.

Thanks for pointing that out. So it appears (as you said) the notebook provided in the course repo does have the correct links to the Rock Paper Scissors dataset and it runs properly…

… if you change this line: from keras_preprocessing.image import ImageDataGenerator
… to this: from tensorflow.keras.preprocessing.image import ImageDataGenerator

But the links in the week 4 instructional materials under the page “Reading:
Introducing the Rock-Paper-Scissors dataset” do not work and need to be updated.

Currently, the page says:


Rock Paper Scissors is a dataset containing 2,892 images of diverse hands in Rock/Paper/Scissors poses. It is licensed CC By 2.0 and available for all purposes, but its intent is primarily for learning and research.

Rock Paper Scissors contains images from a variety of different hands, from different races, ages, and genders, posed into Rock / Paper or Scissors and labeled as such. You can download the training set here, and the test set here. These images have all been generated using CGI techniques as an experiment in determining if a CGI-based dataset can be used for classification against real images. I also generated a few images that you can use for predictions. You can find them here.

Note that all of this data is posed against a white background.

Each image is 300×300 pixels in 24-bit color

You’ll see how this dataset can be used to build a multi-class classifier in the next video.


I’ve reported an issue using the button on the page. I was able to complete both the lab and the programming assignment for the week. Thanks!