reproduce the experimental results

[cripsgreen]

Could you please explain how you trained your model, and how one can reproduce the results reported in your paper, including both the counting and detection tasks?

[jerpelhan]

The training code with experimental setup will be ready in time for the conference. I just had time to release the demo. I’m very sorry for the delay -- I’m currently extremely busy with ongoing research.