The CondInst is not good

[cty-xiaobai]

Hi, I test the CondInst and MaskRcnn in detectron2. The first image is the original image, The second image is the result of maskrcnn, the third image is the result of CondInst .CondInst's results are not good. Do I need to adjust the code？
If anyone has a better solution, welcome to discuss it with me. Thanks a lot!

[tianzhi0549]

@cty-xiaobai I think you should tune the threshold when testing.

[cty-xiaobai]

@cty-xiaobai I think you should tune the threshold when testing.

Thank you for your reply. Which threshold is and how to adjust？

[tianzhi0549]

@cty-xiaobai Please lower the threshold here.

AdelaiDet/demo/demo.py

Line 55 in d459a2f

default=0.5,

[cty-xiaobai]

@cty-xiaobai Please lower the threshold here.

AdelaiDet/demo/demo.py

Line 55 in d459a2f

default=0.5,

OK！thanks for your help！Decrease the threshold, the result is indeed better, but the threshold of detectron2 is also 0.5, I think it should be compared under the same threshold

[tianzhi0549]

@cty-xiaobai please kindly note that different algorithms have different optimal thresholds.

[cty-xiaobai]

@cty-xiaobai please kindly note that different algorithms have different optimal thresholds.

Ok thanks for your help！

[lxtGH]

Hi! @tianzhi0549 what are the better thresholds for demo visualization?