Local Variation as a Statistical Hypothesis Test

TL;DR

This paper analyzes the local variation oversegmentation algorithm, providing a statistical hypothesis testing framework that explains its high performance and introduces a probabilistic variant with state-of-the-art results.

Contribution

It offers a theoretical justification for LV by framing it as a statistical hypothesis test and proposes a probabilistic version with improved accuracy.

Findings

Probabilistic local variation (pLV) achieves state-of-the-art results.

pLV maintains the same computational complexity as LV.

Statistical models explain the success of LV algorithms.

Abstract

The goal of image oversegmentation is to divide an image into several pieces, each of which should ideally be part of an object. One of the simplest and yet most effective oversegmentation algorithms is known as local variation (LV) (Felzenszwalb and Huttenlocher 2004). In this work, we study this algorithm and show that algorithms similar to LV can be devised by applying different statistical models and decisions, thus providing further theoretical justification and a well-founded explanation for the unexpected high performance of the LV approach. Some of these algorithms are based on statistics of natural images and on a hypothesis testing decision; we denote these algorithms probabilistic local variation (pLV). The best pLV algorithm, which relies on censored estimation, presents state-of-the-art results while keeping the same computational complexity of the LV algorithm.