CVJun 11, 2021

Part-aware Panoptic Segmentation

arXiv:2106.06351v173 citations
Originality Incremental advance
AI Analysis

This work defines a novel task for computer vision researchers interested in multi-level scene analysis, though it is incremental as it combines existing segmentation methods.

The paper introduces Part-aware Panoptic Segmentation (PPS), a new scene understanding task that unifies scene parsing and part parsing, and provides annotations on Cityscapes and Pascal VOC datasets along with a Part-aware Panoptic Quality (PartPQ) metric for evaluation.

In this work, we introduce the new scene understanding task of Part-aware Panoptic Segmentation (PPS), which aims to understand a scene at multiple levels of abstraction, and unifies the tasks of scene parsing and part parsing. For this novel task, we provide consistent annotations on two commonly used datasets: Cityscapes and Pascal VOC. Moreover, we present a single metric to evaluate PPS, called Part-aware Panoptic Quality (PartPQ). For this new task, using the metric and annotations, we set multiple baselines by merging results of existing state-of-the-art methods for panoptic segmentation and part segmentation. Finally, we conduct several experiments that evaluate the importance of the different levels of abstraction in this single task.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes