Request for results
Opened this issue · 2 comments
Hi,
Could you please provide original results on different benchmarks?
BR,
Hi, the quick answer is we cannot.
I was visiting another lab last year. As we share all resources in my lab, my desktop has been released and transferred to another guy and he formatted everything. Sad.
I will try to search for the results which might survive someplace on some servers but don't count on that. The majority of data on the servers are some alpha/beta versions of our method.
I will update this issue if I have some founds.
Hi guys, as many people email me about the result, I re-run the experiment with the GitHub version last week and found the result is a little bit different from the paper. I think it is because of the GitHub version has been reconstructed for easy understanding, so they are a little bit different. To reproduce the result, set rotation to 0 in init.m and run the OTB benchmark.
I hope this helps:
OTB result