Guided Stereo Weights don't work

[AnupamYedida]

The PSM Net weights work, but the Guided Stereo weights don't work.

[mattpoggi]

Hi, can you paste here your error? Thanks

[AnupamYedida]

RuntimeError: Error(s) in loading state_dict for DataParallel: Unexpected key(s) in state_dict: "module.feature_extraction.firstconv.0.1.num_batches_tracked", "module.feature_extraction.firstconv.2.1.num_batches_tracked", "module.feature_extraction.firstconv.4.1.num_batches_tracked", "module.feature_extraction.layer1.0.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer1.0.conv2.1.num_batches_tracked", "module.feature_extraction.layer1.1.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer1.1.conv2.1.num_batches_tracked", "module.feature_extraction.layer1.2.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer1.2.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.0.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.0.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.0.downsample.1.num_batches_tracked", "module.feature_extraction.layer2.1.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.1.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.2.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.2.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.3.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.3.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.4.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.4.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.5.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.5.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.6.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.6.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.7.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.7.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.8.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.8.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.9.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.9.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.10.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.10.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.11.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.11.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.12.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.12.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.13.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.13.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.14.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.14.conv2.1.num_batches_tracked", "module.feature_extraction.layer2.15.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer2.15.conv2.1.num_batches_tracked", "module.feature_extraction.layer3.0.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer3.0.conv2.1.num_batches_tracked", "module.feature_extraction.layer3.0.downsample.1.num_batches_tracked", "module.feature_extraction.layer3.1.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer3.1.conv2.1.num_batches_tracked", "module.feature_extraction.layer3.2.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer3.2.conv2.1.num_batches_tracked", "module.feature_extraction.layer4.0.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer4.0.conv2.1.num_batches_tracked", "module.feature_extraction.layer4.1.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer4.1.conv2.1.num_batches_tracked", "module.feature_extraction.layer4.2.conv1.0.1.num_batches_tracked", "module.feature_extraction.layer4.2.conv2.1.num_batches_tracked", "module.feature_extraction.branch1.1.1.num_batches_tracked", "module.feature_extraction.branch2.1.1.num_batches_tracked", "module.feature_extraction.branch3.1.1.num_batches_tracked", "module.feature_extraction.branch4.1.1.num_batches_tracked", "module.feature_extraction.lastconv.0.1.num_batches_tracked", "module.dres0.0.1.num_batches_tracked", "module.dres0.2.1.num_batches_tracked", "module.dres1.0.1.num_batches_tracked", "module.dres1.2.1.num_batches_tracked", "module.dres2.conv1.0.1.num_batches_tracked", "module.dres2.conv2.1.num_batches_tracked", "module.dres2.conv3.0.1.num_batches_tracked", "module.dres2.conv4.0.1.num_batches_tracked", "module.dres2.conv5.1.num_batches_tracked", "module.dres2.conv6.1.num_batches_tracked", "module.dres3.conv1.0.1.num_batches_tracked", "module.dres3.conv2.1.num_batches_tracked", "module.dres3.conv3.0.1.num_batches_tracked", "module.dres3.conv4.0.1.num_batches_tracked", "module.dres3.conv5.1.num_batches_tracked", "module.dres3.conv6.1.num_batches_tracked", "module.dres4.conv1.0.1.num_batches_tracked", "module.dres4.conv2.1.num_batches_tracked", "module.dres4.conv3.0.1.num_batches_tracked", "module.dres4.conv4.0.1.num_batches_tracked", "module.dres4.conv5.1.num_batches_tracked", "module.dres4.conv6.1.num_batches_tracked", "module.classif1.0.1.num_batches_tracked", "module.classif2.0.1.num_batches_tracked", "module.classif3.0.1.num_batches_tracked".

[AnupamYedida]

Also, how do I give SGBM input instead of the files in "disp_occ_0"?