Non-parametric Contextual Relationship Learning for Semantic Video Object Segmentation

TL;DR

This paper introduces a graph-based, non-parametric method for modeling and propagating semantic contextual relationships in videos to improve object segmentation accuracy.

Contribution

It presents a novel exemplar-based, non-parametric approach that encodes relationships on a similarity graph and integrates learned contexts into a CRF for semantic labeling.

Findings

Outperforms state-of-the-art methods on YouTube-Objects dataset

Effectively models spatial-temporal contextual relationships

Enhances semantic segmentation accuracy

Abstract

We propose a novel approach for modeling semantic contextual relationships in videos. This graph-based model enables the learning and propagation of higher-level spatial-temporal contexts to facilitate the semantic labeling of local regions. We introduce an exemplar-based nonparametric view of contextual cues, where the inherent relationships implied by object hypotheses are encoded on a similarity graph of regions. Contextual relationships learning and propagation are performed to estimate the pairwise contexts between all pairs of unlabeled local regions. Our algorithm integrates the learned contexts into a Conditional Random Field (CRF) in the form of pairwise potentials and infers the per-region semantic labels. We evaluate our approach on the challenging YouTube-Objects dataset which shows that the proposed contextual relationship model outperforms the state-of-the-art methods.