LVOS: A Benchmark for Large-scale Long-term Video Object Segmentation

TL;DR

This paper introduces LVOS, a large-scale benchmark dataset with long-duration videos to evaluate and improve long-term video object segmentation models in realistic scenarios.

Contribution

The paper presents LVOS, a new benchmark dataset with over 700 videos averaging 1.14 minutes, designed to better reflect real-world long-term VOS challenges.

Findings

Existing VOS models show significant performance drops on LVOS.

Longer videos increase difficulty in accurate object tracking and segmentation.

LVOS highlights the need for models to handle reappearance and cross-temporal similarities.

Abstract

Video object segmentation (VOS) aims to distinguish and track target objects in a video. Despite the excellent performance achieved by off-the-shell VOS models, existing VOS benchmarks mainly focus on short-term videos lasting about 5 seconds, where objects remain visible most of the time. However, these benchmarks poorly represent practical applications, and the absence of long-term datasets restricts further investigation of VOS in realistic scenarios. Thus, we propose a novel benchmark named LVOS, comprising 720 videos with 296,401 frames and 407,945 high-quality annotations. Videos in LVOS last 1.14 minutes on average, approximately 5 times longer than videos in existing datasets. Each video includes various attributes, especially challenges deriving from the wild, such as long-term reappearing and cross-temporal similar objects. Compared to previous benchmarks, our LVOS better reflects VOS models' performance in real scenarios. Based on LVOS, we evaluate 20 existing VOS models under 4 different settings and conduct a comprehensive analysis. On LVOS, these models suffer a large performance drop, highlighting the challenge of achieving precise tracking and segmentation in real-world scenarios. Attribute-based analysis indicates that key factor to accuracy decline is the increased video length, emphasizing LVOS's crucial role. We hope our LVOS can advance development of VOS in real scenes. Data and code are available at https://lingyihongfd.github.io/lvos.github.io/.