Image Parsing with a Wide Range of Classes and Scene-Level Context

TL;DR

This paper introduces a nonparametric scene parsing method that enhances accuracy and class coverage by combining probabilistic classifiers and global semantic context, achieving state-of-the-art results on large datasets.

Contribution

It presents a novel nonparametric approach that integrates multiple classifiers and global label costs without relying on image retrieval, improving scene parsing performance.

Findings

Achieved state-of-the-art accuracy on SIFTflow dataset.

Obtained near-record results on LMSun dataset.

Enhanced class coverage and representation of less-represented classes.

Abstract

This paper presents a nonparametric scene parsing approach that improves the overall accuracy, as well as the coverage of foreground classes in scene images. We first improve the label likelihood estimates at superpixels by merging likelihood scores from different probabilistic classifiers. This boosts the classification performance and enriches the representation of less-represented classes. Our second contribution consists of incorporating semantic context in the parsing process through global label costs. Our method does not rely on image retrieval sets but rather assigns a global likelihood estimate to each label, which is plugged into the overall energy function. We evaluate our system on two large-scale datasets, SIFTflow and LMSun. We achieve state-of-the-art performance on the SIFTflow dataset and near-record results on LMSun.