Open anustupdas opened 5 years ago
Sorry for late reply. It is semi-supervised VOS, which means an initial object mask is essential. If you have an initial mask, applying our method to streaming video is straightforward as it can process video frame-by-frame.
Thank you for the implementation.
Can you provide any idea how I can test this on real time video ?