<- Back to Datasets
LVIS v0.5
A DATASET FOR LARGE VOCABULARY INSTANCE SEGMENTATION
LVIS is a new dataset for long tail object instance segmentation. 1000+ Categories: found by data-driven object discovery in 164k images Long Tail: category discovery naturally reveals a large number of rare categories Masks: more than 2.2 million high quality instance segmentation masks
Facebook AI Research
View author website
Task
Object Detection
Annotation Types
Bounding Boxes
164000
Items
1000
Classes
164000
Labels
Models using this dataset
Last updated on
January 20, 2022
Licensed under
Research Only
