sunghoonim / DPSNet

[ICLR19] DPSNet: End-to-end Deep Plane Sweep Stereo

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Reproduce the benchmark

lawsonsli opened this issue · comments

I cannot reproduce the benchmark using your code and pretrained model.

dataset method Abs_rel Abs_diff Sq_rel rms Log_rms a1 a2 a3
mvs Dpsnet-paper 0.0722 0.2095 0.0798 0.4928 0.1527 0.8930 0.9502 0.9760
mvs Dpsnet-code 0.0809 0.1901 0.0660 0.3996 0.1531 0.8952 0.9580 0.9851
sun3d Dpsnet-paper 0.1470 0.3234 0.1071 0.4269 0.1906 0.7892 0.9317 0.9672
sun3d Dpsnet-code 0.1576 0.3404 0.1268 0.4538 0.1998 0.7917 0.9341 0.9790

This table is what I got by running test.py on sun3d and mvs. Can you help me about this?

Thanks sincerely for your extraordinary work and code.