📚 Documentation Issue
Which config corresponds to the centernet * proposed in the paper? Is there a config for this model? You mentioned in your paper that you were inspired by GFL to improve centernet to become centernet *, but didn't GFL improve focal loss? Does it seem that there is no GFL loss in the code? Is there a specific formula for the Loss function used by centernet *? Thank you.
📚 Documentation Issue
Which config corresponds to the centernet * proposed in the paper? Is there a config for this model? You mentioned in your paper that you were inspired by GFL to improve centernet to become centernet *, but didn't GFL improve focal loss? Does it seem that there is no GFL loss in the code? Is there a specific formula for the Loss function used by centernet *? Thank you.