Multiple Object Track and Segmentation
1 papers with code • 1 benchmarks • 1 datasets
Multiple object tracking and segmentation requires detecting, tracking, and segmenting objects belonging to a set of given classes.
(Image and definition credit: Prototypical Cross-Attention Networks for Multiple Object Tracking and Segmentation, NeurIPS 2021, Spotlight )
Benchmarks
These leaderboards are used to track progress in Multiple Object Track and Segmentation
| Trend | Dataset | Best Model | Paper | Code | Compare |
|---|---|---|---|---|---|
|
BDD100K
Most implemented papers
Prototypical Cross-Attention Networks for Multiple Object Tracking and Segmentation
We propose Prototypical Cross-Attention Network (PCAN), capable of leveraging rich spatio-temporal information for online multiple object tracking and segmentation.
