Closed feifeiwei closed 4 months ago
Thanks your attention to this work. The results presented in this method, as well as in other compare methods are obtained by training on three combined datasets. Therefore, I provide only one weight for these three datasets. Training on independent datasets may result in some metrics being raised or lowered.
Thanks, i will have a try!
"Thank you for your nice work. The result I obtained from training on the IRSTD-1K dataset is different from that in the paper. Do you have any suggestions to help reproduce it? thanks!
here is my result:
IRSTD-1K_SCTransNet_Fri_Jun__7_21_33_28_2024.txt