A question on duplicated data in BB dataset about multi_part_assembly

[Wuziyi616]

We indeed observed a few cases like this. However, they are not duplicates -- the geometry of these meshes has some very small differences (it is hard to tell these minor differences from the point cloud as it somewhat blurs the cutting interface), and thus it's hard to de-duplicate them. Also, I guess such issues are much less frequent in the more-part cases, because the fracturing pattern is much more diverse.

[wzm2256]

We indeed observed a few cases like this. However, they are not duplicates -- the geometry of these meshes has some very small differences (it is hard to tell these minor differences from the point cloud as it somewhat blurs the cutting interface), and thus it's hard to de-duplicate them. Also, I guess such issues are much less frequent in the more-part cases, because the fracturing pattern is much more diverse.

That is not 'a few' cases. As I said, most of the data are highly similar, if not duplicated, at least in the 2 parts case. This should be noted in the dataset for clarification. Anyway, since you have already close this issue, I won't complain about this further.

[Wuziyi616]

I said "a few" meaning it doesn't happen a lot in the entire dataset. Indeed, the 2-part data may have many such issues as 2 is so limited in the way it fractured. I'm not the geometry simulation person in the project, so I cannot comment more on this issue.