Closed DD-BONED closed 2 years ago
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。
Let me know if this question is now addressed.
但是经过我打印发现len_c - unionk有些是非0的这个怎么理解
但是经过我打印发现len_c - unionk有些是非0的这个怎么理解
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).
如果咋样,为什么还是要坚持是giou呢?
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.
Mark as resolved.
你好你使用了giouloss但是在这个任务中,c和(AuB)不应该是相等的吗