61 lines
1.9 KiB
Python
61 lines
1.9 KiB
Python
import glob
|
|
import os.path as osp
|
|
|
|
from dassl.utils import listdir_nohidden
|
|
|
|
from ..build import DATASET_REGISTRY
|
|
from ..base_dataset import Datum, DatasetBase
|
|
|
|
|
|
@DATASET_REGISTRY.register()
|
|
class VLCS(DatasetBase):
|
|
"""VLCS.
|
|
|
|
Statistics:
|
|
- 4 domains: CALTECH, LABELME, PASCAL, SUN
|
|
- 5 categories: bird, car, chair, dog, and person.
|
|
|
|
Reference:
|
|
- Torralba and Efros. Unbiased look at dataset bias. CVPR 2011.
|
|
"""
|
|
|
|
dataset_dir = "VLCS"
|
|
domains = ["caltech", "labelme", "pascal", "sun"]
|
|
data_url = "https://drive.google.com/uc?id=1r0WL5DDqKfSPp9E3tRENwHaXNs1olLZd"
|
|
|
|
def __init__(self, cfg):
|
|
root = osp.abspath(osp.expanduser(cfg.DATASET.ROOT))
|
|
self.dataset_dir = osp.join(root, self.dataset_dir)
|
|
|
|
if not osp.exists(self.dataset_dir):
|
|
dst = osp.join(root, "vlcs.zip")
|
|
self.download_data(self.data_url, dst, from_gdrive=True)
|
|
|
|
self.check_input_domains(
|
|
cfg.DATASET.SOURCE_DOMAINS, cfg.DATASET.TARGET_DOMAINS
|
|
)
|
|
|
|
train = self._read_data(cfg.DATASET.SOURCE_DOMAINS, "train")
|
|
val = self._read_data(cfg.DATASET.SOURCE_DOMAINS, "crossval")
|
|
test = self._read_data(cfg.DATASET.TARGET_DOMAINS, "test")
|
|
|
|
super().__init__(train_x=train, val=val, test=test)
|
|
|
|
def _read_data(self, input_domains, split):
|
|
items = []
|
|
|
|
for domain, dname in enumerate(input_domains):
|
|
dname = dname.upper()
|
|
path = osp.join(self.dataset_dir, dname, split)
|
|
folders = listdir_nohidden(path)
|
|
folders.sort()
|
|
|
|
for label, folder in enumerate(folders):
|
|
impaths = glob.glob(osp.join(path, folder, "*.jpg"))
|
|
|
|
for impath in impaths:
|
|
item = Datum(impath=impath, label=label, domain=domain)
|
|
items.append(item)
|
|
|
|
return items
|