helioszhao / shade-visualdg

Style-Hallucinated Dual Consistency Learning: A Unified Framework for Visual Domain Generalization

Hi there, I'm Yuyang Zhao 👋

Contact Me:

✉️ Email: [email protected]

🔗 Website: https://yuyangzhao.com

🔎 Google Scholar: https://scholar.google.com/citations?user=u5M6XPAAAAAJ

shade-visualdg's Issues

How to use our custom dataset in object detection?

Thanks

Visualization of style-diversified images

Hi, in the MiT backbone, the diverse image features are generated by shm

SHADE-VisualDG/sseg/mmseg/models/backbones/mix_transformer.py

Line 425 in 954bfc4

x_tuple = self.shm(x)

How can I visualize the style-diversified images from x_tuple, is there any related code for visualizing the images?

Thanks for your help.

A basic problem about the Fig. 4 in paper SHADE

Hello, I'm interested in the work on Style-Hallucinated and I want to creat visualizations similar to "Fig. 4: Visualization of distributions of different domains" in your paper. I'd like to analyze the distributions of various datasets. Could you please provide details on the method used to generate those visualizations?
Thanks for your help！