Yichong Lu

Results 3 issues of Yichong Lu

你好,很感谢你的工作,不过看了你的代码,感觉有几处地方和原文有出入:1.原文中训练和推理阶段的3D bounding box是用3D detector得到的,但代码中似乎只在pre-processing中用了2D detector得到mask,并没有用3D detector得到3D bounding box的步骤 2. 原文的encoder似乎是两个独立开来的,代码中的encoder似乎只是一个。不知道这些出入是不是确实存在,会不会影响最后的效果?

The live demo on huggingface seems to have some errors. Will you fix it soon?

Thanks for the nice work! I noticed that you have updated a new checkpoint named Wonder3D plus. However, I don't find the technical details about this checkpoint, such as how...