WaymoPerceptionDatasetProcessor¶
- class standard_e2e.caching.src_datasets.waymo_perception.WaymoPerceptionDatasetProcessor(common_output_path, split, index_data_generator=None, adapters=None, context_aggregators=None)[source]¶
Bases:
SourceDatasetProcessorProcessor for the Waymo Perception dataset.
HD-map handling: Waymo’s proto puts
map_featuresonly on the first frame of each segment. This processor caches the decoded world-frame map features persegment_idand applies the per-frame inverse pose on every frame.WaymoPerceptionDatasetConverterpre-populates the cache before parallel work begins, so both sequential and parallel preprocessing produce HD maps for every frame.- Parameters:
common_output_path (str)
split (str)
index_data_generator (IndexDataGenerator | None)
adapters (list[AbstractAdapter] | None)
context_aggregators (list[SegmentContextAggregator] | None)
- CAMERAS_ORDER = {CameraDirection.FRONT: 1, CameraDirection.FRONT_LEFT: 2, CameraDirection.FRONT_RIGHT: 3, CameraDirection.SIDE_LEFT: 4, CameraDirection.SIDE_RIGHT: 5}¶
- DATASET_NAME = 'waymo_perception'¶
- property allowed_splits: list[str]¶
Return the list of allowed splits for the Waymo Perception dataset.
- property context_aggregators¶
- needs_attr(attr)¶
Whether at least one registered adapter reads this
StandardFrameDatafield. Used by per-dataset processors to skip expensive modality builds (cameras, lidar, hd_map, detections, …) when no adapter would consume them.Truewhenattris in the consumed-attrs union, plus a hard-coded special case: the identifier / index fields are always treated as needed since they are required for the cache + index regardless of adapter chain.- Return type:
- Parameters:
attr (StandardFrameDataField)