This dataset contains each ShapeNet object rendered from 25 random perspectives on transparent background. Images are rendered at 512x512 and then cropped to fit the object, so they vary in size:
ds = tfds.load("shapenet_pretraining", data_dir="gs://kubric-public/tfds") The dataset contains the following information:
- "image_id":
str - "asset_id":
str
The id of the ShapeNet object. E.g. "02691156/1021a0914a7207aff927ed529ad90a11". - "label":
int
One of the 55 Shapenet classes: ["airplane", "ashcan", "bag", "basket", "bathtub", "bed", "bench", "birdhouse", "bookshelf", "bottle", "bowl", "bus", "cabinet", "camera", "can", "cap", "car", "cellular telephone", "chair", "clock", "computer keyboard", "dishwasher", "display", "earphone", "faucet", "file", "guitar", "helmet", "jar", "knife", "lamp", "laptop", "loudspeaker", "mailbox", "microphone", "microwave", "motorcycle", "mug", "piano", "pillow", "pistol", "pot", "printer", "remote control", "rifle", "rocket", "skateboard", "sofa", "stove", "table", "telephone", "tower", "train", "vessel", "washer"] - "camera_position":
(3,) [float32]position of the camera in a half-sphere shell with inner radius 9 and outer radius 10. The object sits at the origin. - "image":
(None, None, 4) [uint8]The rendered image in RGBA format, cropped to fit the object.
Generate single scene with the shapenet_pretraining_worker.py script:
docker run --rm --interactive \
--user $(id -u):$(id -g) \
--volume "$(pwd):/kubric" \
kubricdockerhub/kubruntu \
/usr/bin/python3 challenges/shapenet_pretraining/shapenet_pretraining_worker.py See shapenet_pretraining.py for the TFDS definition / conversion.