I am training for the dataset PASCAL VOC with ResNet 152. I have removed last layer of weights (_detectron mask rcnn resnet 152 model) which and used as Pretrained model. after training converges, If i run a demo, i am getting following image
Mask is drawn next to Bounding box.
Is this issue on Demo code or Training itself is not proper?
I am training for the dataset PASCAL VOC with ResNet 152. I have removed last layer of weights (_detectron mask rcnn resnet 152 model) which and used as Pretrained model. after training converges, If i run a demo, i am getting following image
Mask is drawn next to Bounding box.
Is this issue on Demo code or Training itself is not proper?
How can i debug it? Where is the problem?