RandomResizedCrop in PyTorch (3)

This content originally appeared on DEV Community and was authored by Super Kai (Kazuya Ito)

*Memos:

My post explains RandomResizedCrop() about size argument.
My post explains RandomResizedCrop() about scale argument.
My post explains OxfordIIITPet().

RandomResizedCrop() can crop a random part of an image, then resize it to a given size as shown below:

from torchvision.datasets import OxfordIIITPet
from torchvision.transforms.v2 import RandomResizedCrop
from torchvision.transforms.functional import InterpolationMode

origin_data = OxfordIIITPet(
    root="data",
    transform=None
)

s1000r1_1origin_data = OxfordIIITPet( # `s` is size and `r` is ratio.
    root="data",
    transform=RandomResizedCrop(size=1000)
)

s1000r01_10_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.1, 10])
)

s1000r01_1_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.1, 1])
)

s1000r1_10_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[1, 10])
)

s1000r09_09_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.9, 0.9])
)

s1000r08_08_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.8, 0.8])
)

s1000r07_07_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.7, 0.7])
)

s1000r06_06_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.6, 0.6])
)

s1000r05_05_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.5, 0.5])
)

s1000r04_04_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.4, 0.4])
)

s1000r03_03_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.3, 0.3])
)

s1000r02_02_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.2, 0.2])
)

s1000r01_01_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.1, 0.1])
)

s1000r001_001_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.01, 0.01])
)

s1000r0001_0001_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.001, 0.001])
)

s1000r00001_00001_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[0.0001, 0.0001])
)

s1000r2_2_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[2, 2])
)

s1000r3_3_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[3, 3])
)

s1000r4_4_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[4, 4])
)

s1000r5_5_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[5, 5])
)

s1000r6_6_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[6, 6])
)

s1000r7_7_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[7, 7])
)

s1000r8_8_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[8, 8])
)

s1000r9_9_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[9, 9])
)

s1000r10_10_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[10, 10])
)

s1000r100_100_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[100, 100])
)

s1000r1000_1000_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[1000, 1000])
)

s1000r10000_10000_data = OxfordIIITPet(
    root="data",
    transform=RandomResizedCrop(size=1000, ratio=[10000, 10000])
)

import matplotlib.pyplot as plt

def show_images1(data, main_title=None):
    plt.figure(figsize=[10, 5])
    plt.suptitle(t=main_title, y=0.8, fontsize=14)
    for i, (im, _) in zip(range(1, 6), data):
        plt.subplot(1, 5, i)
        plt.imshow(X=im)
    plt.tight_layout()
    plt.show()

show_images1(data=origin_data, main_title="origin_data")
print()
show_images1(data=s1000r1_1origin_data, main_title="s1000r1_1origin_data")
show_images1(data=s1000r01_10_data, main_title="s1000r01_10_data")
show_images1(data=s1000r01_1_data, main_title="s1000r01_1_data")
show_images1(data=s1000r1_10_data, main_title="s1000r1_10_data")
print()
show_images1(data=s1000r1_1origin_data, main_title="s1000r1_1origin_data")
show_images1(data=s1000r09_09_data , main_title="s1000r09_09_data ")
show_images1(data=s1000r08_08_data, main_title="s1000r08_08_data")
show_images1(data=s1000r07_07_data, main_title="s1000r07_07_data")
show_images1(data=s1000r06_06_data, main_title="s1000r06_06_data")
show_images1(data=s1000r05_05_data, main_title="s1000r05_05_data")
show_images1(data=s1000r04_04_data, main_title="s1000r04_04_data")
show_images1(data=s1000r03_03_data, main_title="s1000r03_03_data")
show_images1(data=s1000r02_02_data, main_title="s1000r02_02_data")
show_images1(data=s1000r01_01_data, main_title="s1000r01_01_data")
show_images1(data=s1000r001_001_data, main_title="s1000r001_001_data")
show_images1(data=s1000r0001_0001_data, main_title="s1000r0001_0001_data")
show_images1(data=s1000r00001_00001_data, main_title="s1000r00001_00001_data")
print()
show_images1(data=s1000r1_1origin_data, main_title="s1000r1_1origin_data")
show_images1(data=s1000r2_2_data, main_title="s1000r2_2_data")
show_images1(data=s1000r3_3_data, main_title="s1000r3_3_data")
show_images1(data=s1000r4_4_data, main_title="s1000r4_4_data")
show_images1(data=s1000r5_5_data, main_title="s1000r5_5_data")
show_images1(data=s1000r6_6_data, main_title="s1000r6_6_data")
show_images1(data=s1000r7_7_data, main_title="s1000r7_7_data")
show_images1(data=s1000r8_8_data, main_title="s1000r8_8_data")
show_images1(data=s1000r9_9_data, main_title="s1000r9_9_data")
show_images1(data=s1000r10_10_data, main_title="s1000r10_10_data")
show_images1(data=s1000r100_100_data, main_title="s1000r100_100_data")
show_images1(data=s1000r1000_1000_data, main_title="s1000r1000_1000_data")
show_images1(data=s1000r10000_10000_data, main_title="s1000r10000_10000_data")

# ↓ ↓ ↓ ↓ ↓ ↓ The code below is identical to the code above. ↓ ↓ ↓ ↓ ↓ ↓ 
def show_images2(data, main_title=None, s=None, sc=(0.08, 1.0),
                 r=(0.75, 1.3333333333333333),
                 ip=InterpolationMode.BILINEAR, a=True):
    plt.figure(figsize=[10, 5])
    plt.suptitle(t=main_title, y=0.8, fontsize=14)
    for i, (im, _) in zip(range(1, 6), data):
        plt.subplot(1, 5, i)
        if s:
            rrc = RandomResizedCrop(size=s, scale=sc, # Here
                                    ratio=r, interpolation=ip,
                                    antialias=a)
            plt.imshow(X=rrc(im)) # Here
        else:
            plt.imshow(X=im)
    plt.tight_layout()
    plt.show()

show_images2(data=origin_data, main_title="origin_data")
print()
show_images2(data=origin_data, main_title="s1000r1_1origin_data", s=1000,  
             r=[1, 1])
show_images2(data=origin_data, main_title="s1000r01_10_data", s=1000,
             r=[0.1, 10])
show_images2(data=origin_data, main_title="s1000r01_1_data", s=1000,
             r=[0.1, 1])
show_images2(data=origin_data, main_title="s1000r1_10_data", s=1000, r=[1, 10])
print()
show_images2(data=origin_data, main_title="s1000r1_1origin_data", s=1000,
             r=[1, 1])
show_images2(data=origin_data, main_title="s1000r09_09_data", s=1000,
             r=[0.9, 0.9])
show_images2(data=origin_data, main_title="s1000r08_08_data", s=1000,
             r=[0.8, 0.8])
show_images2(data=origin_data, main_title="s1000r07_07_data", s=1000,
             r=[0.7, 0.7])
show_images2(data=origin_data, main_title="s1000r06_06_data", s=1000,
             r=[0.6, 0.6])
show_images2(data=origin_data, main_title="s1000r05_05_data", s=1000,
             r=[0.5, 0.5])
show_images2(data=origin_data, main_title="s1000r04_04_data", s=1000,
             r=[0.4, 0.4])
show_images2(data=origin_data, main_title="s1000r03_03_data", s=1000,
             r=[0.3, 0.3])
show_images2(data=origin_data, main_title="s1000r02_02_data", s=1000,
             r=[0.2, 0.2])
show_images2(data=origin_data, main_title="s1000r01_01_data", s=1000,
             r=[0.1, 0.1])
show_images2(data=origin_data, main_title="s1000r001_001_data", s=1000,
             r=[0.01, 0.01])
show_images2(data=origin_data, main_title="s1000r0001_0001_data", s=1000,
             r=[0.001, 0.001])
show_images2(data=origin_data, main_title="s1000r00001_00001_data", s=1000,
             r=[0.0001, 0.0001])
print()
show_images2(data=origin_data, main_title="s1000r1_1origin_data", s=1000,
             r=[1, 1])
show_images2(data=origin_data, main_title="s1000r2_2_data", s=1000, r=[2, 2])
show_images2(data=origin_data, main_title="s1000r3_3_data", s=1000, r=[3, 3])
show_images2(data=origin_data, main_title="s1000r4_4_data", s=1000, r=[4, 4])
show_images2(data=origin_data, main_title="s1000r5_5_data", s=1000, r=[5, 5])
show_images2(data=origin_data, main_title="s1000r6_6_data", s=1000, r=[6, 6])
show_images2(data=origin_data, main_title="s1000r7_7_data", s=1000, r=[7, 7])
show_images2(data=origin_data, main_title="s1000r8_8_data", s=1000, r=[8, 8])
show_images2(data=origin_data, main_title="s1000r9_9_data", s=1000, r=[9, 9])
show_images2(data=origin_data, main_title="s1000r10_10_data", s=1000,
             r=[10, 10])
show_images2(data=origin_data, main_title="s1000r100_100_data", s=1000,
             r=[100, 100])
show_images2(data=origin_data, main_title="s1000r1000_1000_data", s=1000,
             r=[1000, 1000])
show_images2(data=origin_data, main_title="s1000r10000_10000_data", s=1000,
             r=[10000, 10000])

This content originally appeared on DEV Community and was authored by Super Kai (Kazuya Ito)

Print Share Comment Cite Upload Translate Updates

APA

Super Kai (Kazuya Ito) | Sciencx (2025-02-13T01:56:58+00:00) RandomResizedCrop in PyTorch (3). Retrieved from https://www.scien.cx/2025/02/13/randomresizedcrop-in-pytorch-3/

MLA

" » RandomResizedCrop in PyTorch (3)." Super Kai (Kazuya Ito) | Sciencx - Thursday February 13, 2025, https://www.scien.cx/2025/02/13/randomresizedcrop-in-pytorch-3/

HARVARD

Super Kai (Kazuya Ito) | Sciencx Thursday February 13, 2025 » RandomResizedCrop in PyTorch (3)., viewed ,<https://www.scien.cx/2025/02/13/randomresizedcrop-in-pytorch-3/>

VANCOUVER

Super Kai (Kazuya Ito) | Sciencx - » RandomResizedCrop in PyTorch (3). [Internet]. [Accessed ]. Available from: https://www.scien.cx/2025/02/13/randomresizedcrop-in-pytorch-3/

CHICAGO

" » RandomResizedCrop in PyTorch (3)." Super Kai (Kazuya Ito) | Sciencx - Accessed . https://www.scien.cx/2025/02/13/randomresizedcrop-in-pytorch-3/

IEEE

" » RandomResizedCrop in PyTorch (3)." Super Kai (Kazuya Ito) | Sciencx [Online]. Available: https://www.scien.cx/2025/02/13/randomresizedcrop-in-pytorch-3/. [Accessed: ]

rf:citation

» RandomResizedCrop in PyTorch (3) | Super Kai (Kazuya Ito) | Sciencx | https://www.scien.cx/2025/02/13/randomresizedcrop-in-pytorch-3/ |

Please log in to upload a file.

There are no updates yet.
Click the Upload button above to add an update.

You must be logged in to translate posts. Please log in or register.

Related Posts