Closed FUTUREEEEEE closed 3 years ago
the test results looks very good but the output imgs are all black (most of the output img are small ,about 2k each)
should results looks like this? or i need to post-process them?
Hi @FUTUREEEEEE , Here is a function for visualization. https://github.com/YudeWang/semantic-segmentation-codebase/blob/995b8fa9d6b08e3ff1f15a7aef2ce5445deaa834/lib/datasets/VOCDataset.py#L115
the test results looks very good but the output imgs are all black (most of the output img are small ,about 2k each)
should results looks like this? or i need to post-process them?