what is the difference between enabling/disabling this? would the bbox be outside of the ROI otherwise?
If you enable this, we will resize the bbox that exceeds the image size accordingly. If you disable this, we will just remove those boxes.
why would the network be seeing pixels outside of the roi defined in the preprocessed tensor inputs?
It’s not for the preprocessed data, it’s for the postprocessed data. The inference from some models may output values that are out of image.
There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.