A series of basic algorithms that are useful for video understanding, including Single Object Tracking (SOT), Video Object Segmentation (VOS) and so on.
I'm searching for the implementation of 'state score' which conisists of a 'confidence score' and a ' concentration score', as stated in《State-Aware Tracker for Real-Time Video Object Segmentation》. But it seems that this part is not included in current repository . Can you tell me the path to that or will it be added some day? Especially, I wonder how the 'concentration score' is calculated.
I'm searching for the implementation of 'state score' which conisists of a 'confidence score' and a ' concentration score', as stated in《State-Aware Tracker for Real-Time Video Object Segmentation》. But it seems that this part is not included in current repository . Can you tell me the path to that or will it be added some day? Especially, I wonder how the 'concentration score' is calculated.