poincarelee

I can declare that you didn't search and examine the code carefully. Actually I used resnet101 these days, and the same problems still exist: 

hi， I use flownet2's predict results as groundtruth, since in real scene groundtruth for optical flow couldn't be obtained. As for div_flow, yes, you are right. Any other tricks I...

Ok, thanks for your timely reply. 1) good advice, I will try RAFT's prediction ASAP. 2) flow ground truth divided by 20 has already been taken into the code; I...

I have tried replacing IRR-PWC with Fastflownet directly and correct some bugs, however, it didn't work. It seems that the datasets and somewhere in loss.py needs to be modified.

I trained HMDB51 dataset for 20 epochs with modelA0_stream_statedict_v3, the result is as follows: 

@papasanimohansrinivas , any progress in reproducing of streaming A3, A4, A5 ? looking eagerly forward to your work.

Thanks, Could you kindly tell me which parameters are required ? Seems parameters needed in the fitting command are all specified. OpenGL has not been installed indeed, I will install...


 This is the picture I used, morevrer the json file of the keypoints generated by running openpose are put in the corresponding folder. However , I didnot overlay the...

I see. Single person in an image would be better? I will try.