AI Paper Atlas

Research map / Video / Video segmentation and tracking

LVOS: A Benchmark for Large-Scale Long-Term Video Object Segmentation

Lingyi Hong, Zhongying Liu, Wenchao Chen, Chenzhi Tan, Yuang Feng, Xinyu Zhou, Pinxue Guo, Jinglun Li, Zhaoyu Chen, Shuyong Gao, Wei Zhang, Wenqiang Zhang

TPAMI 2025 · Video › Video segmentation and tracking · approach: Matching-based video object segmentation · 41 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add LVOS training videos to training dataDeAOT-B · LVOS val semi-supervisedJ&F: 63.3 -> 67.3Improved
add LVOS training videos to training dataDeAOT-L · LVOS val semi-supervisedJ&F: 63.9 -> 67.1Improved
add LVOS training videos to training dataXMem · LVOS val semi-supervisedJ&F: 64.5 -> 67.4Improved
add LVOS training videos to training dataAFB-URR · LVOS val semi-supervisedJ&F: 44.1 -> 49.8Improved
add LVOS training videos to training dataRDE · LVOS val semi-supervisedJ&F: 62.2 -> 65.1Improved
add LVOS training videos to training dataLWL · LVOS val semi-supervisedJ&F: 60.6 -> 61.4Improved
add LVOS training videos to training dataCFBI · LVOS val semi-supervisedJ&F: 55.0 -> 54.6No clear effect
add LVOS training videos to training dataLWL · LVOS test semi-supervisedJ&F: 60.9 -> 59.8Hurt
add LVOS training videos to training dataSimulFlow · LVOS val unsupervised singleJ&F: 42.7 -> 42.2No clear effect
add LVOS training videos to training dataTMO · LVOS test unsupervised singleJ&F: 45.9 -> 52.8Improved
restrict object location with oracle box (ablation)DeAOT-L · LVOS val oracle (semi-supervised)J&F: 67.1 -> 78.4Improved
restrict object location with oracle box (ablation)XMem · LVOS val oracle (semi-supervised)J&F: 73.3 -> 68.5Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy