Looks like high/xhigh both get the back leg z-index incorrect relative to the back wheel of the bike and frame.
The frame/leg problem was discussed here previously and likely trickled into the training set.
The problem you pointed out was never discussed on hacker news, and it clearly shows they still lack spatial reasoning. I wish you didn’t mention it. :)
The frame/leg problem was discussed here previously and likely trickled into the training set.
The problem you pointed out was never discussed on hacker news, and it clearly shows they still lack spatial reasoning. I wish you didn’t mention it. :)