NiuKeke comments

Results 16 comments of


                                            NiuKeke

Hello man, why did you extract the features from the octave of the smallest resolution instead of original resolution?

Ok,i will send it. wait a moment.

Hello man, why did you extract the features from the octave of the smallest resolution instead of original resolution?

> @NiuKeke Are you discussing the ExtractSiftOctave function which currently using constant scale to extract features? > Could you please share your code by commit it to github or make...

Hello man, why did you extract the features from the octave of the smallest resolution instead of original resolution?

> The reason is to make it as scale-invariant as possible. From the theory, the gradients for the descriptor should be computed at a scale that corresponds to the scale...

Hello man, why did you extract the features from the octave of the smallest resolution instead of original resolution?

> To me, it doesn't seem like a good idea to use the original resolution for the descriptor. Large features, detected at a coarse scale, will then contain high-frequency components...

How to estimate the accuracy of converting HIP to CUDA or converting CUDA to HIP?

> @NiuKeke I'm not sure we're clear on what kind of accuracy you mean. Could you elaborate? I mean if there were some test suites about CUDA/HIP just like opencl-cts,...

How to estimate the accuracy of converting HIP to CUDA or converting CUDA to HIP?

> This could be done by some test suites about fp operations. Unfortunately, I am not aware of one for CUDA/HIP. > > There is a research paper about math...