SFD
SFD copied to clipboard
LittlePey
Sparse Fuse Dense: Towards High Quality 3D Detection with Depth Completion (CVPR 2022, Oral)
Hi, I have a question. I replaced giou in the configuration file with diou, and the code can be trained normally without encountering the Nan situation, which may be because...
您好，不好意思打扰一下，有个问题想咨询一下。伪点云支路中，有如下操作：1.扩大roi的边框，然后选取其中的伪点云；2.查询伪点云中各点相邻的9个近邻点（根据（u，v）查询）。我的问题是roi不是很准确的情况下，其roi内部的伪点云可能是来自不同场景（由于gt增强、旋转等），如果同一个roi中的这些来自不同场景中的伪点坐标（u，v）存在重复，应该是会对近邻的查询产生干扰吧。不知道代码中有没有什么手段可以避免伪点云的坐标（u，v）可能存在重复的问题。谢谢。
Hello, your work is very meaningful, but when I reproduced this work, I verified that the results in the paper could be achieved on the validation set. However, after submitting...
Hello author, Recently, I have developed an object detection model and trained it on 3D KITTI dataset. Now, I would like to test on the test set and generate the...
May I ask you how the pseudo point cloud in Figure 1 of the paper is visualized？ 
Hello author! I want to ask a question: In order to submit to KITTI and obtain better test results When dividing training set and validation set If I use 80%-20%...
Always reporting errors on different samples, Has anyone encountered the same problem before？  
 亲爱的作者你好，请问为什么对于这些检测的代码，公布出来的pth文件我直接放入test中，得出的结果都是比公布的结果低的？公布的结果是我图像中标出的R40_3D,0.7，0.7,0.7这一行的结果吗？此外我还想问都是R40_0.7,为什么Car [email protected], 0.70, 0.70:Car [email protected], 0.50, 0.50:这两个的第一列结果是不一样的,参数不都是一样吗？感谢您花时间阅读
Hi authors, appreciations for your exellent work and code contribution I once was curious about that how the model's performance could degrade when abnormal images are given. Thus I tried...
Great work, how are the files under "depth_dense_twise" and "depth_pseudo_rgbseguv_twise" folders generated? Can I get the file corresponding to the KITTI dataset tracking task image?
