The Multidimensional Wisdom of Crowds

TLDR

Distributing labeling tasks among hundreds or thousands of annotators is an increasingly important method for annotating large datasets. The authors propose a method to estimate the true class of each image from noisy annotations supplied by multiple annotators. Their approach models image formation and annotation in an abstract Euclidean space, treating each annotator as a multidimensional entity with competence, expertise, and bias, enabling the discovery of annotator and image groups with distinct skill sets and characteristics. The model outperforms state‑of‑the‑art methods on synthetic and real data, and, starting from binary labels, reveals rich information such as annotator schools of thought and distinct image categories.

Abstract

Distributing labeling tasks among hundreds or thousands of annotators is an increasingly important method for annotating large datasets. We present a method for estimating the underlying value (e.g. the class) of each image from (noisy) annotations provided by multiple annotators. Our method is based on a model of the image formation and annotation process. Each image has different characteristics that are represented in an abstract Euclidean space. Each annotator is modeled as a multidimensional entity with variables representing competence, expertise and bias. This allows the model to discover and represent groups of annotators that have different sets of skills and knowledge, as well as groups of images that differ qualitatively. We find that our model predicts ground truth labels on both synthetic and real data more accurately than state of the art methods. Experiments also show that our model, starting from a set of binary labels, may discover rich information, such as different schools of thought amongst the annotators, and can group together images belonging to separate categories.