jucic

question about the performance on my own data

1

Thanks for your nice job. I tried the pretrained model on my own data, however the disparity result looks not good because of the blurry depth edge and incorrect depth...

第七章3d2d运行时报错（用的新版的g2o但是都改成了unique_ptr所以不是这个指针的问题）

6

./pose_estimation_3d2d 1.png 2.png 1_depth.png 2_depth.png *** Error in `./pose_estimation_3d2d': double free or corruption (out): 0x0000000000d931e0 *** ======= Backtrace: ========= /lib/x86_64-linux-gnu/libc.so.6(+0x777e5)[0x7fc37d73d7e5] /lib/x86_64-linux-gnu/libc.so.6(+0x8037a)[0x7fc37d74637a] /lib/x86_64-linux-gnu/libc.so.6(cfree+0x4c)[0x7fc37d74a53c] /usr/local/lib/libg2o_core.so(_ZN3g2o7Factory12registerTypeERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEEPNS_32AbstractHyperGraphElementCreatorE+0x497)[0x7fc3805e3667]

about autoflow

1

thanks for your nice job, will you release the code of autoflow or the pretrained weights of pwc-net trained by autoflow dataset?

Flow result is a mess when the adjcent two frames are almost the same(no action or very small action).

5

 look at the above picture, when the adjcent two frames are almost the same(no action or very small action), Flow result is a mess, I guess this is because...

about the performance on middlebury dataset.

1

   hello, thanks for your nice job. I test the finetuning_model on some middlebury images, however, in some cases the performance is not satisfying. do you know the...

thanks for your nice job, will you release the code of autoflow or the pretrained weights of pwc-net trained by autoflow dataset?

Thanks for your wonderful job. I tried to remove the "+1" inmake_cost_volume_v2, and the predicted disparity of the full image turns out to be a litte smaller than before, and...