loss about actionformer_release HOT 8 CLOSED

你说的对，实质上GIoU Loss在1D，并且中心点为同一个的时候，会退化成IoU Loss。

from actionformer_release.

Let me know if this question is now addressed.

from actionformer_release.

但是经过我打印发现len_c - unionk有些是非0的这个怎么理解

from actionformer_release.

但是经过我打印发现len_c - unionk有些是非0的这个怎么理解

from actionformer_release.

If you print the actual values of len_c - unionk, they are either zero or very close to zero (due to limited floating point precision).

from actionformer_release.

如果咋样，为什么还是要坚持是giou呢？

from actionformer_release.

With our current implementation, GIoU is indeed equivalent to IoU. One can either call our regression loss as GIoU or IoU. This equivalence, however, can not be generalized to the temporal action localization task, and is only due to the design of our model outputs. The GIoU loss has been proven to be more effective than IoU loss for regressing action boundaries. Should our design evolves, GIoU is preferred over IoU.

from actionformer_release.

Mark as resolved.

from actionformer_release.