D_Kohei

icon indicating a website link https://www.cs.cit.tum.de/sse/dozono/

icon location Ph.D in security testing @ Technical University of Munich (TUM) | M.Sc. Informatics @ TUM

Results 4 comments of D_Kohei

About accuracy

Same here, I couldn't reach to the accuracy mentioned on the README...

类别不均衡

Thank you for reaching me. Do you speak English?

[Feature] inference_recognizer supports more than video & dict input types

I have the same problem. In mmaction2 ver 0.24.1, we can pass RawFrame data. However, the latest version of mmaction2 only allows us to pass "video" data...

Why the report-data only contains the code covrage data?

Hi there and @Ice1187 . Is there any possible way of making a bug targeting benchmark dataset manually?