TAP-Vid is a benchmark which contains both real-world videos with accurate human annotations of point tracks, and synthetic videos with perfect ground-truth point tracks. This is designed for a new task called tracking any point.
Variants: TAP-Vid
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| Point Tracking | PIPs++ | PointOdyssey: A Large-Scale Synthetic Dataset … | 2023-07-27 |
Recent papers with results on this dataset: