Open xyl-507 opened 2 weeks ago
Regarding the log backup, it seems that the original logs may not have been retained due to the time elapsed. However, you could reproduce the results by re-running the experiments and evaluating the tracking metrics across different time points and training epochs, which should be relatively straightforward.
Thanks again for your understanding!
Dear author @hoqolo : Thank you very much for your work.
I noticed in Figure 1 of the supplemental material that you compared the training process. Could you please share the training logs of SDSTrack and ViPT?
Thanks in advance!