--- license: mit pretty_name: LEAD 123D viewer: false tags: - autonomous-driving - carla - py123d - end-to-end-driving - multimodal --- # LEAD 123D A privileged expert drives CARLA Leaderboard 2.0 routes and writes every modality it sees to the standardized [Py123D](https://github.com/kesai-labs/py123d) format: cameras, lidar, radar, the full scene state and the maps to place it on. Code, loaders, visualizers and closed-loop benchmarks live in [kesai-labs/lead](https://github.com/kesai-labs/lead). Policy learning is one application. The logs carry the whole scene, so perception, world models, video generation and self-play RL read the same files. ## Contents | Path | Size | Logs | | ----------------------- | ------ | ----- | | `logs/normal_view` | 651 GB | 8,800 | | `logs/perturbated_view` | 416 GB | 8,800 | | `maps/carla` | 789 MB | — | `logs////` holds one route, across 43 scenario types. `perturbated_view` is the same routes with the camera rig rotated and translated, which is what lets a policy learn to recover from its own drift. `maps/carla` holds the 12 CARLA towns, Town01 through Town15. The training cache the LEAD repository can read is **not** part of this dataset. It is derived from these logs and policy-specific; `scripts/common/build_cache.sh` recomputes it. ## Modalities Each log directory holds 32 Arrow files: - **Cameras**, six positions (`pcam_f0`, `pcam_b0`, `pcam_l0`, `pcam_l1`, `pcam_r0`, `pcam_r1`) × RGB, depth, semantic and instance segmentation - **Lidar** (`lidar.lidar_top`) and **radar** (`radar.radar_merged`) - **Scene state**: ego state, 3D box detections, traffic light detections - **Route** (`route_position`): derived from the ego odometry, not from the planner: the polyline is the path the ego actually drove, so the ego sits on the route at every frame, extended past the end of the log by the expert's planner route. Per-frame progress along it is stored. Records what it was computed from (needs py123d >= 0.7) - **Metadata**: `custom.driving_meta`, `sync` Point clouds and scene state are written at the native **20 Hz** tick rate. Camera streams are written on every fifth tick (**4 Hz**) to keep storage tractable. ## Download The dataset is roughly a terabyte over half a million files, so fetch the part you need. `logs/normal_view` and `maps` alone are enough to train and evaluate. ```bash pip install "huggingface-hub[hf_xet]" hf download ln2697/lead-123d --repo-type dataset --local-dir lead-123d \ --include 'logs/normal_view/*' 'maps/*' 'config.yaml' ``` Downloads resume, so an interrupted one is continued by running it again. Drop `--include` for everything, perturbated views included. ## Usage The logs are plain Py123D, so the standard API reads them without LEAD: ```python from py123d.api.scene.arrow.arrow_scene_builder import ArrowSceneBuilder from py123d.api.scene.scene_filter import SceneFilter from py123d.common.execution.thread_pool_executor import ThreadPoolExecutor # Enumeration is lazy; nothing is decoded here. scenes = ArrowSceneBuilder( logs_root="lead-123d/logs", maps_root="lead-123d/maps", ).get_scenes( SceneFilter( split_names=["normal_view"], future_num_iterations=40, # 2 s at the 20 Hz tick rate required_scene_modalities=["camera:all@initial"], # drop partial rigs ), ThreadPoolExecutor(), ) # Modalities are read per frame, by iteration index. ego = scenes[0].get_ego_state_se3_at_iteration(0) # EgoStateSE3 boxes = scenes[0].get_box_detections_se3_at_iteration(0) # BoxDetectionsSE3 ``` To train on it, point `PY123D_DATA_ROOT` at the download and follow the [LEAD README](https://github.com/kesai-labs/lead). ## Citation ```bibtex @inproceedings{Nguyen2026CVPR, author = {Long Nguyen and Micha Fauth and Bernhard Jaeger and Daniel Dauner and Maximilian Igl and Andreas Geiger and Kashyap Chitta}, title = {LEAD: Minimizing Learner-Expert Asymmetry in End-to-End Driving}, booktitle = {Conference on Computer Vision and Pattern Recognition (CVPR)}, year = {2026}, } @article{Dauner2026ARXIV, author = {Dauner, Daniel and Charraut, Valentin and Berle, Bastian and Li, Tianyu and Nguyen, Long and Wang, Jiabao and Jing, Changhui and Igl, Maximilian and Caesar, Holger and Ivanovic, Boris and Geiger, Andreas and Chitta, Kashyap}, title = {123D: Unifying Multi-Modal Autonomous Driving Data at Scale}, journal = {arXiv preprint arXiv:2605.08084}, year = {2026}, } ```