README.md

---
license: other
license_name: upstream-assets-see-acknowledgements
pretty_name: munich2drive
task_categories:
- video-classification
tags:
- autonomous-driving
- trajectory-prediction
- motion-planning
- carla
- commonroad
- sumo
- surround-view
- munich
size_categories:
- n<1K
dataset_info:
  features:
  - name: mp4
    dtype: video
  - name: clip_id
    dtype: string
  - name: scenario_id
    dtype: int64
  - name: ego_id
    dtype: int64
  - name: version
    dtype: string
  - name: map
    dtype: string
  - name: weather
    dtype: string
  - name: sun_altitude_deg
    dtype: float64
  - name: precipitation
    dtype: float64
  - name: wetness
    dtype: float64
  - name: night
    dtype: bool
  - name: n_agents
    dtype: int64
  - name: agents_by_type
    struct:
    - name: CAR
      dtype: int64
    - name: TRUCK
      dtype: int64
    - name: BUS
      dtype: int64
    - name: BICYCLE
      dtype: int64
    - name: MOTORCYCLE
      dtype: int64
  - name: horizon_steps
    dtype: int64
  - name: dt_s
    dtype: float64
  - name: audit_clean
    dtype: bool
  - name: max_pose_error_m
    dtype: float64
  - name: max_height_dev_m
    dtype: float64
  - name: ground_layer_corrections
    dtype: int64
  - name: actors_inherited
    dtype: int64
  - name: actors_surviving_cleanup
    dtype: int64
  - name: ego_blueprint
    dtype: string
  - name: ego_collision_events
    dtype: int64
  - name: ego_collisions
    list:
    - name: time_step
      dtype: int64
    - name: other
      dtype: string
    - name: other_id
      dtype: int64
    - name: impulse
      dtype: float64
  - name: trajectory
    struct:
    - name: t
      list: int64
    - name: x
      list: float64
    - name: y
      list: float64
    - name: carla_x
      list: float64
    - name: carla_y
      list: float64
    - name: yaw
      list: float64
    - name: v
      list: float64
    - name: a
      list: float64
    - name: yaw_rate
      list: float64
  - name: n_states
    dtype: int64
  - name: path_length_m
    dtype: float64
  - name: speed_mps
    struct:
    - name: mean
      dtype: float64
    - name: max
      dtype: float64
  - name: ego_length_m
    dtype: float64
  - name: ego_width_m
    dtype: float64
  - name: trajectory_source
    dtype: string
  splits:
  - name: train
    num_bytes: 5646498356
    num_examples: 500
  download_size: 5646727502
  dataset_size: 5646498356
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

# munich2drive

Surround-camera driving clips from a **real Munich HD map** built by
[SimForge](https://simforge.ai/), each paired with the **ego vehicle's full trajectory**. Every clip is a nine-panel video -- six surround cameras at
the ego's real sensor extrinsics, a chase view, a bird's-eye view, and the frame-locked
CommonRoad ground truth -- so what the model sees and what the data says are visible side
by side.

- **500 clips** over 50 base scenarios (500 sumo)
- 61,200 agent-replays, 79-161 agents per clip
- 12 weather / time-of-day conditions, 20 at night
- ego paths 10.64-232.85 m (median 108.96 m)

## Each row

| field | |
|---|---|
| `mp4` | the 3x3 grid clip (1920x1080, 10 fps, 120 frames) |
| `json` | everything below, as one record |
| `trajectory` | parallel arrays: `t`, `x`, `y`, `carla_x`, `carla_y`, `yaw`, `v`, `a`, `yaw_rate` |
| `path_length_m`, `speed_mps` | trajectory summary |
| `weather`, `sun_altitude_deg`, `precipitation`, `wetness`, `night` | conditions |
| `n_agents`, `agents_by_type` | surrounding traffic |
| `ego_collision_events`, `ego_collisions` | ego collision-sensor events with impulses |
| `audit_clean`, `max_pose_error_m`, `max_height_dev_m` | per-clip verification, below |

`x`/`y` are the CommonRoad map frame -- the frame the CommonRoad panel of the video is
drawn in. `carla_x`/`carla_y` are the same states in the CARLA world frame, via the
verified transform (`carla_x = lx`, `carla_y = -ly` after reprojecting EPSG:3857 to the
map's local transverse-Mercator frame; `crdesigner` emits 3857, which is ~1.5x
scale-distorted at 48N, so this conversion is not optional).

The trajectory is given in world coordinates; ego-relative future waypoints follow from
the state at any `t` by the usual rigid transform.

## Verification

Rendering audits each clip against its scenario file at 14 time steps and ships the
result, so these are measurements, not claims:

- `audit_clean`: **500 / 500**
- agent pose error vs CommonRoad: median **0.003 m**, p99 0.022 m, max 1.297 m
- height deviation: median 0.034 m, p99 0.248 m, max 0.302 m
- clips rendered with actors inherited from a previous run: **0**; with actors surviving cleanup: **0**
- clips with at least one ego collision event: 6 of 500. Of the 11 events, **7 are
  contacts with static traffic signs** and 4 with vehicles -- the sensor reports contact
  with anything, so this is not a vehicle-collision rate

## Known limitations

- Traffic is replayed kinematically from recorded SUMO trajectories and does **not** react
  to the ego. The ego collisions present are therefore predominantly rear-end contacts
  caused by non-reactive traffic, not ego-at-fault conflicts.
- Where SUMO placed two footprints overlapping, the render shows them interpenetrating.
- The ego is rendered with a stand-in mesh (`vehicle.volkswagen.t2_2021`), identical
  across all clips. Its *footprint* throughout the CommonRoad and SUMO data is the
  real EDGAR size, 4.973 x 1.941 x 2.320 m, so the trajectories are unaffected.

## Acknowledgements

Munich HD map from [SimForge](https://simforge.ai/); ego vehicle and motion planner from
[TUM-AVS](https://github.com/TUM-AVS) (EDGAR digital twin, Frenetix); built on
[CARLA](https://carla.org), [CommonRoad](https://commonroad.in.tum.de) and
[SUMO](https://eclipse.dev/sumo/). Observe the upstream licences of the HD map, simulator
assets and vehicle models.