Dataset card: variants
Browse files
README.md
CHANGED
|
@@ -25,7 +25,7 @@ measures the reaction of driving models by comparing the trajectory they predict
|
|
| 25 |
|
| 26 |
## Files
|
| 27 |
|
| 28 |
-
- `reactnav/<object>_tau<τ>s.tar.gz`: the 12
|
| 29 |
look-ahead time τ = 0.5, 1.0, 1.5 or 2.0 s. Each holds the camera frames the inserted object changes and
|
| 30 |
one metadata file per sample describing the object. The unchanged frames are the original navtest images.
|
| 31 |
- `reactnav/lanes.tar`: the NAVSIM lane polygons the ReactNAV scorer uses.
|
|
@@ -34,7 +34,7 @@ measures the reaction of driving models by comparing the trajectory they predict
|
|
| 34 |
|
| 35 |
ReactNAV builds on your copy of NAVSIM navtest. Follow
|
| 36 |
[Getting the data](https://github.com/niktheod/reactbench-ad#getting-the-data) in the code repository: extract the
|
| 37 |
-
|
| 38 |
`python -m reactbench.datasets.reactnav link` adds the unchanged frames from navtest.
|
| 39 |
|
| 40 |
```bash
|
|
|
|
| 25 |
|
| 26 |
## Files
|
| 27 |
|
| 28 |
+
- `reactnav/<object>_tau<τ>s.tar.gz`: the 12 edited variants, a `crossing_pedestrian`, `standing_pedestrian` or `car` at
|
| 29 |
look-ahead time τ = 0.5, 1.0, 1.5 or 2.0 s. Each holds the camera frames the inserted object changes and
|
| 30 |
one metadata file per sample describing the object. The unchanged frames are the original navtest images.
|
| 31 |
- `reactnav/lanes.tar`: the NAVSIM lane polygons the ReactNAV scorer uses.
|
|
|
|
| 34 |
|
| 35 |
ReactNAV builds on your copy of NAVSIM navtest. Follow
|
| 36 |
[Getting the data](https://github.com/niktheod/reactbench-ad#getting-the-data) in the code repository: extract the
|
| 37 |
+
variant archives into `datasets/reactnav/` and `lanes.tar` into `outputs/work/navsim/`, then
|
| 38 |
`python -m reactbench.datasets.reactnav link` adds the unchanged frames from navtest.
|
| 39 |
|
| 40 |
```bash
|