Visualizations on cross-attention weight in different decoder stages

zeliu98 / Group-Free-3D

Group-Free 3D Object Detection via Transformers

MIT License

243 stars 33 forks source link

Visualizations on cross-attention weight in different decoder stages #3

Closed jlqzzz closed 3 years ago

jlqzzz commented 3 years ago

Dear author, Thank you for your good work. I want to know how to visualize the results in Figure 5. Can you provide the corresponding visualization code?

zeliu98 commented 3 years ago

Hi, I use open3d to visualize it, by setting different colors to points according to the attention weight.

abhigoku10 commented 2 years ago

@zeliu98 can you please share ur code for visualization it will be helpful Thanks in advance