Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
File size: 1,230 Bytes
19a285e 495efc8 19a285e 495efc8 19a285e 495efc8 19a285e b2997cc 19a285e c3f97b9 19a285e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 |
import gradio as gr
import kornia as K
from kornia.core import concatenate, Tensor
def rescale_aa(file, height, width):
img: Tensor = K.io.load_image(file.name, K.io.ImageLoadType.RGB32)
img = img[None]
img_rescale: Tensor = K.geometry.rescale(img, (float(height),float(width)),antialias=False)
img_rescale_aa: Tensor = K.geometry.rescale(img, (float(height),float(width)),antialias=True)
img_out = concatenate([img_rescale, img_rescale_aa], -1)
# when antialiasing , some values are going greater than 1 i.e 1.00001 which is giving error while displaying the output image,so clipping the output values from 0 to 1
return K.utils.tensor_to_image(img_out.clamp_(0, 1))
examples = [
["examples/a.png",0.25,0.25],
["examples/iron_man.jpeg",0.25,0.25],
]
kornia_resizing_demo = gr.Interface(
rescale_aa,
[
gr.inputs.Image(type="file"),
gr.inputs.Slider(minimum=0.005, maximum=2, step=0.005, default=0.25, label="Height"),
gr.inputs.Slider(minimum=0.005, maximum=2, step=0.005, default=0.25, label="Width")
],
"image",
examples=examples,
live=False,
enable_queue = True,
allow_flagging = "never"
)
kornia_resizing_demo.launch() |