Pascal-Part

The 'Pascal-Part' dataset is a benchmark that contains annotated images for evaluating open-vocabulary part segmentation by providing fine-grained part labels and complex boundaries.

Papers using Pascal-Part (3)

PartSeg: Few-shot Part Segmentation via Part-aware Prompt Learning2023 · 2 citesObjects as context for detecting their semantic parts2017 · 1 citesCompositor: Bottom-up Clustering and Compositing for Robust Part and Object Segmentation2023