This repository provides images, data and code for organizing images into clusters using the crowd - see our arXiv report here.

$ cd src
$ python run_orchestra.py scenes/intelligent_samples --categorization

This will output precision, recall, accuracy and number of unclustered items for the scenes dataset. In addition, you'll also see visualization of confusion matrix in data/scenes/intelligent_samples/categorization.

See python run_orchestra.py -h for an explanation of the command line options.