search

Linfeng-Tang / SeAFusion

The code of " Image fusion in the loop of high-level vision tasks: A semantic-aware real-time infrared and visible image fusion network"
MIT License
184 stars 34 forks source link

About Label #6

Open 870572761 opened 1 year ago

870572761 commented 1 year ago

Why the dataset's labels are all black?

Linfeng-Tang commented 1 year ago

In the original labels, 0, 1, 2, 3, ... , 8 are used to represent the different categories, so all the labels are black. You can write a mapping code to visualize the labels.

870572761 commented 1 year ago

In the original labels, 0, 1, 2, 3, ... , 8 are used to represent the different categories, so all the labels are black. You can write a mapping code to visualize the labels.

Thanks.

870572761 commented 1 year ago

I find some labels just exist one type. Is it a special label or my wrong? There is my code. import numpy as np from PIL import Image import torch label_path1='.\leir\label\js.jpg' a=Image.open(label_path1) label = np.array(a)

label = np.asarray(Image.fromarray(label), dtype=np.int64)

lb=torch.tensor(label) lb = np.uint8(255.0 * lb)

lb = Image.fromarray(lb) lb.show() `

Linfeng-Tang commented 1 year ago

Apologies for the delay in responding. Have you resolved the previous issues or do you still need assistance? If you still need help, I can share the mapping code I have recently written with you.

870572761 commented 1 year ago

Thanks.I really need that.It is my email: l870572761@gmail.com.

Apologies for the delay in responding. Have you resolved the previous issues or do you still need assistance? If you still need help, I can share the mapping code I have recently written with you.

Thanks.I really need that.It is my email: l870572761@gmail.com.

870572761 commented 1 year ago

Apologies for the delay in responding. Have you resolved the previous issues or do you still need assistance? If you still need help, I can share the mapping code I have recently written with you.

谢谢作者，我非常需要这个映射代码，我的邮箱是：l870572761@gmail.com