A Benchmark Dataset and Evaluation Methodology for Video Object Segmentation

TLDR

Datasets and benchmarks are essential for progress in computer vision, yet legacy sets can hinder advancement by saturating performance and lacking contemporary, high‑quality data. This work introduces a new benchmark dataset and evaluation methodology for video object segmentation. The DAVIS dataset contains fifty Full HD videos with pixel‑accurate, per‑frame ground truth, covering challenges such as occlusion, motion blur, and appearance changes, and the authors evaluate state‑of‑the‑art methods using spatial extent, silhouette accuracy, and temporal coherence metrics. The analysis highlights the strengths and weaknesses of current approaches, suggesting promising directions for future research.

Abstract

Over the years, datasets and benchmarks have proven their fundamental importance in computer vision research, enabling targeted progress and objective comparisons in many fields. At the same time, legacy datasets may impend the evolution of a field due to saturated algorithm performance and the lack of contemporary, high quality data. In this work we present a new benchmark dataset and evaluation methodology for the area of video object segmentation. The dataset, named DAVIS (Densely Annotated VIdeo Segmentation), consists of fifty high quality, Full HD video sequences, spanning multiple occurrences of common video object segmentation challenges such as occlusions, motionblur and appearance changes. Each video is accompanied by densely annotated, pixel-accurate and per-frame ground truth segmentation. In addition, we provide a comprehensive analysis of several state-of-the-art segmentation approaches using three complementary metrics that measure the spatial extent of the segmentation, the accuracy of the silhouette contours and the temporal coherence. The results uncover strengths and weaknesses of current approaches, opening up promising directions for future works.