Can you give me some visualization suggestions?

[mcd-1021]

Thank you for your work. Currently, I also want to conduct visual analysis of the running results. I want to know the meaning of “val_probs” in https://github.com/LiyaoTang/ERDA/blob/819c4c671be2169634dd1d59507de3340864f154/utils/tester.py#L179 and how it corresponds to each point in the point cloud

Besides,I have added the following code at the end of this code to visulaize the result

write_ply(test_name, [points, val_probs], ['x', 'y', 'z', 'preds'])

That is to say, I don't know which variable the "points" in the above code should be written as. Can you give me some suggestions? @LiyaoTang

[LiyaoTang]

Hi,

Thanks for your interest in our work!

The val_probs is a list of arrays, and each array corresponds to a cloud/scene in the dataset. The list is in the same order as the dataset.input_trees[split], where the split='validation' here.

Besides, please note that the points in input_trees are subsampled from the original scene. That is why we pass the projections there.

Best, Liyao

[mcd-1021]

Thank you for your prompt reply. I have made some attempts. When I run the following code：

write_ply("12345.ply",
                        [dataset.input_trees['validation'], val_probs],
                        ['x', 'y', 'z', 'preds'])

I encountered an error： 'list' object has no attribute 'ndim' on line 253 of the ply. py if field.ndim < 2: Is it a problem with this variable “dataset.input_trees['validation']” ？ How should I solve this problem？Can you give me some advice？

[lrnb12345]

感谢您的及时答复。 我已经做了一些尝试。 当我运行以下代码时：

write_ply("12345.ply",
                        [dataset.input_trees['validation'], val_probs],
                        ['x', 'y', 'z', 'preds'])

我遇到了一个错误： 'list' object has no attribute 'ndim' on line 253 of the ply. py if field.ndim < 2: 是不是这个变量“dataset.input_trees['validation']”有问题？ 我该如何解决这个问题？能给我一些建议吗？

Hello, may I ask if your problem has been resolved