Problem about runtime of base model

[fangchuan]

Hi, thank you for sharing the awesome work. I have run the basemodel raftstereo-middlebury.pth without any refinement on my costume dataset. The precision of results are pretty good, But the runtime of model prediction seems does not match that described in the article. I add some time-analysis snippet in demo.py:

The command predicts stereo on custome data is : python demo.py --restore_ckpt models/raftstereo-middlebury.pth -l=output_xvisio/rect_cam0/*.jpg -r=output_xvisio/rect_cam1/*.jpg --corr_implementation alt --mixed_precision

The configuration of my local machine:

• image resolution: 640 x 400;
• CPU: Intel® Core™ i7-8700K CPU @ 3.70GHz × 12
• GPU: NVIDIA GeForce GTX 1080 Runtime of base model:

The configuration of my server machine:

• image resolution: 640 x 400;
• CPU:Intel(R) Xeon(R) CPU E5-2682 v4 @ 2.50GHz
• GPU: NVIDIA Tesla P100 Runtime of base model:

Could you help me to figure out what mistake I made caused this problem? I really appreciate your help!

[lahavlipson]

--corr_implementation alt saves memory but is slower, although you shouldn't need it for individual 640 x 400 resolution images. Timings in the paper were measured using --corr_implementation reg_cuda, using an NVIDIA RTX-2080ti

[fangchuan]

--corr_implementation alt saves memory but is slower, although you shouldn't need it for individual 640 x 400 resolution images. Timings in the paper were measured using --corr_implementation reg_cuda, using an NVIDIA RTX-2080ti

thank your reply. I follow your advice to adjust the --corr_implementation reg_cuda command, the runtime of raftstereo-middlebury.pth on Tesla P100 is about 400ms.

Another problem is about model raftstereo-realtime.pth, I load this model but consistently run into Errors when loading state_dict, the github page gives no clue about how to run the realtime model, can you help me?

[lahavlipson]

Are you able to run python demo.py --restore_ckpt models/raftstereo-realtime.pth --shared_backbone --n_downsample 3 --n_gru_layers 2 --slow_fast_gru --valid_iters 7 --corr_implementation reg_cuda --mixed_precision from the README? If not, can you post the error message?

[lahavlipson]

Closing this issue due to inactivity. Feel free to reopen if you still have problems or questions.