closedaxis-12573/partial_map_semantic_localization_dataset_gazebo

Map image contains the objects on which the model has been trained, but the scene has other objects too. --- dataset_info: features:

• —name: case_id dtype: string
• —name: scene_id dtype: int32
• —name: camera dtype: image
• —name: lidar dtype: image
• —name: map dtype: image
• —name: pose dtype: string
• —name: grid dtype: string
• —name: orientation dtype: string
• —name: valid_cell dtype: bool
• —name: x_raw dtype: int64
• —name: y_raw dtype: int64
• —name: x_norm dtype: float64
• —name: y_norm dtype: float64
• —name: theta_deg dtype: float64
• —name: theta_rad dtype: float64
• —name: dir_cos dtype: float64
• —name: dir_sin dtype: float64 splits:
• —name: train numbytes: 1607877954 numexamples: 5990
• —name: test numbytes: 402103697 numexamples: 1498 downloadsize: 1659595765 datasetsize: 2009981651 configs:
• —configname: default datafiles:
• —split: train path: data/train-*
• —split: test path: data/test-* ---