slee5777 / dcnet Goto Github PK

Dear @slee5777 and @bearpelican,

thank you for making the work publicly available, including the dataset and the code.

I have a question about Figure 4 in the paper, where you computed AP across different classes. It is not clear from the paper what was the training and what testing dataset to generate the results for Figure 4. The holdout dataset is only mentioned that was used to generate results for Figure 5, where only cell counts are reported.

Can you elaborate on that, so that Figure 4 results for the DCNet can be reproduced?

Thanks!