paper-with-me

LVIS

홈페이지 · 논문 551편

LVIS is a dataset for long tail instance segmentation. It has annotations for over 1000 object categories in 164k images. Source: LVIS

Images

벤치마크

Open Vocabulary Object Detection on LVIS v1.0 결과 168개

Object Detection on LVIS v1.0 val 결과 75개

Zero-Shot Object Detection on LVIS v1.0 minival 결과 55개

Zero-Shot Object Detection on LVIS v1.0 val 결과 45개

Few-Shot Object Detection on LVIS v1.0 val 결과 42개

Few-Shot Object Detection on LVIS v1.0 test-dev 결과 30개

Object Detection on LVIS v1.0 minival 결과 30개

Instance Segmentation on LVIS v1.0 val 결과 25개

Novel Object Detection on LVIS v1.0 val 결과 5개

Object Detection on LVIS v1.0 결과 5개

Instance Segmentation on LVIS v1.0 test-dev 결과 1개

Long-tailed Object Detection on LVIS v1.0 val 결과 1개