Set threshold resnet18 pytorch

from PIL import Image
import torch
from torchvision import transforms
input_image = Image.open("img/danbooru_resnet1.png") # load an image of your choice
preprocess = transforms.Compose([
    transforms.Resize(360),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.7137, 0.6628, 0.6519], std=[0.2970, 0.3017, 0.2979]),
])
input_tensor = preprocess(input_image)
input_batch = input_tensor.unsqueeze(0) # create a mini-batch as expected by the model

if torch.cuda.is_available():
    input_batch = input_batch.to('cuda')
    model.to('cuda')

with torch.no_grad():
    output = model(input_batch)

# The output has unnormalized scores. To get probabilities, you can run a sigmoid on it.
probs = torch.sigmoid(output[0]) # Tensor of shape 6000, with confidence scores over Danbooru's top 6000 tags

# Second part of the code 
# to set a threshold value and to use it 
tmp = probs[probs > thresh]
inds = probs.argsort(descending=True)

Source: github.com

Similar codes

0 vote264 views1 answer
How to compress image field in django?
Python Django
0 vote220 views2 answers
Python suppress print output from function
Python
1 vote236 views1 answer
Python swapcase() method Example
Python
0 vote213 views1 answer
Reg.fit(x_train, y_train) TypeError: fit() missing 1 required positional argument: 'y'
Python
0 vote226 views1 answer
Beautifulsoup remove tag with class
Python
1 vote214 views2 answers
Python long
Python
1 vote214 views1 answer
Python list remove at index
Python
58 votes316 views6 answers
Python class definition
Python
142 votes243 views6 answers
While loop python
Python
10 votes241 views2 answers
Rotation turtle python
Python