HongChuang/ultralytics-ascend
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 3

ultralytics 8.2.52 fix CenterCrop transforms for PIL Image inputs (#14308)

Browse source 
Signed-off-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: UltralyticsAssistant <web@ultralytics.com>
Co-authored-by: Lucas Buligon Antunes <lukasbuligonantunes@gmail.com>
This commit is contained in:
Glenn Jocher 2024-07-10 03:00:14 +02:00 • committed by GitHub
parent 755dcd6ca0
commit 997f2c92cd
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
20 changed files with 38 additions and 30 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ultralytics/data/augment.py
View file

@ -1401,6 +1401,8 @@ class CenterCrop:
Returns:
(numpy.ndarray): The center-cropped and resized image as a numpy array.
"""
if isinstance(im, Image.Image): # convert from PIL to numpy array if required
im = np.asarray(im)
imh, imw = im.shape[:2]
m = min(imh, imw) # min dimension
top, left = (imh - m) // 2, (imw - m) // 2

14
ultralytics/data/dataset.py
View file

@ -15,6 +15,7 @@ from torch.utils.data import ConcatDataset
from ultralytics.utils import LOCAL_RANK, NUM_THREADS, TQDM, colorstr
from ultralytics.utils.ops import resample_segments
from ultralytics.utils.torch_utils import TORCH_1_13
from .augment import (
Compose,
@ -263,7 +264,7 @@ class YOLOMultiModalDataset(YOLODataset):
super().__init__(*args, data=data, task=task, **kwargs)
def update_labels_info(self, label):
"""Add texts information for multi modal model training."""
"""Add texts information for multi-modal model training."""
labels = super().update_labels_info(label)
# NOTE: some categories are concatenated with its synonyms by `/`.
labels["texts"] = [v.split("/") for _, v in self.data["names"].items()]
@ -296,10 +297,10 @@ class GroundingDataset(YOLODataset):
with open(self.json_file, "r") as f:
annotations = json.load(f)
images = {f'{x["id"]:d}': x for x in annotations["images"]}
imgToAnns = defaultdict(list)
img_to_anns = defaultdict(list)
for ann in annotations["annotations"]:
imgToAnns[ann["image_id"]].append(ann)
for img_id, anns in TQDM(imgToAnns.items(), desc=f"Reading annotations {self.json_file}"):
img_to_anns[ann["image_id"]].append(ann)
for img_id, anns in TQDM(img_to_anns.items(), desc=f"Reading annotations {self.json_file}"):
img = images[f"{img_id:d}"]
h, w, f = img["height"], img["width"], img["file_name"]
im_file = Path(self.img_path) / f
@ -416,7 +417,10 @@ class ClassificationDataset:
import torchvision # scope for faster 'import ultralytics'
# Base class assigned as attribute rather than used as base class to allow for scoping slow torchvision import
self.base = torchvision.datasets.ImageFolder(root=root, allow_empty=True)
if TORCH_1_13: # 'allow_empty' argument first introduced in torch 1.13
self.base = torchvision.datasets.ImageFolder(root=root, allow_empty=True)
else:
self.base = torchvision.datasets.ImageFolder(root=root)
self.samples = self.base.samples
self.root = self.base.root