Below are links to supplemental materials to our paper, Learning Visual Object Categories for Robot Affordance Prediction.

Code for PeopleBot experiments
This archive includes the code needed to run our first set of experiments and the numeric data extracted from each image in the data set for those experiments.

Code for Rovio experiments
This includes the code for the second set of experiments using the Rovio. Greyscale images from our data set are included, as well as the numeric feature data extracted from each image.

Rovio dataset
This has the greyscale images from the data set used for the Rovio experiments.

Footage of Rovio experiment
This is a video summary of our experiment with the Rovio as described in the paper. Note: the DivX codec, available for free at, is required to view this video.