search

JiangWenPL / multiperson

Code repository for the paper: "Coherent Reconstruction of Multiple Humans from a Single Image" in CVPR'20
https://jiangwenpl.github.io/multiperson/
376 stars 50 forks source link

evaluation #39

Closed HuiliCui closed 2 years ago

HuiliCui commented 2 years ago

hi，when i run python3 tools/full_eval.py configs/smpl/tune.py haggling --ckpt ./work_dirs/tune/latest.pth to evaluation Panoptic, there is something wrong, 2022-03-05 16:17:40,458 - WARNING - unexpected key in source state_dict: fc.weight, fc.bias

missing keys in source state_dict: layer3.4.bn3.num_batches_tracked, layer3.5.bn3.num_batches_tracked, layer4.0.downsample.1.num_batches_tracked, layer1.0.downsample.1.num_batches_tracked, layer1.2.bn1.num_batches_tracked, layer2.2.bn2.num_batches_tracked, layer1.1.bn2.num_batches_tracked, layer2.3.bn1.num_batches_tracked, layer4.2.bn2.num_batches_tracked, layer3.5.bn1.num_batches_tracked, layer1.2.bn2.num_batches_tracked, layer3.5.bn2.num_batches_tracked, layer3.0.downsample.1.num_batches_tracked, layer3.2.bn2.num_batches_tracked, layer1.0.bn2.num_batches_tracked, layer2.1.bn1.num_batches_tracked, layer3.3.bn1.num_batches_tracked, layer4.2.bn3.num_batches_tracked, layer3.1.bn3.num_batches_tracked, layer3.3.bn3.num_batches_tracked, layer4.1.bn2.num_batches_tracked, layer2.3.bn3.num_batches_tracked, layer3.2.bn1.num_batches_tracked, layer4.2.bn1.num_batches_tracked, layer2.2.bn1.num_batches_tracked, layer3.0.bn2.num_batches_tracked, layer4.0.bn3.num_batches_tracked, layer2.2.bn3.num_batches_tracked, layer2.0.bn1.num_batches_tracked, layer2.3.bn2.num_batches_tracked, layer1.0.bn1.num_batches_tracked, layer2.1.bn2.num_batches_tracked, layer2.0.downsample.1.num_batches_tracked, layer3.3.bn2.num_batches_tracked, layer3.1.bn2.num_batches_tracked, layer3.0.bn1.num_batches_tracked, layer3.4.bn1.num_batches_tracked, layer4.1.bn3.num_batches_tracked, layer4.0.bn1.num_batches_tracked, layer2.0.bn3.num_batches_tracked, layer3.1.bn1.num_batches_tracked, layer1.1.bn3.num_batches_tracked, layer3.2.bn3.num_batches_tracked, layer1.0.bn3.num_batches_tracked, layer3.4.bn2.num_batches_tracked, layer2.0.bn2.num_batches_tracked, layer2.1.bn3.num_batches_tracked, layer4.0.bn2.num_batches_tracked, layer1.1.bn1.num_batches_tracked, layer4.1.bn1.num_batches_tracked, layer1.2.bn3.num_batches_tracked, layer3.0.bn3.num_batches_tracked, bn1.num_batches_tracked

Traceback (most recent call last): File "tools/full_eval.py", line 297, in main() File "tools/full_eval.py", line 222, in main train_dataset = get_dataset(cfg.datasets[0].train) File "/home/cuihuili/coherent_reconstruction/multiperson-master/mmdetection/mmdet/datasets/utils.py", line 111, in get_dataset dset = obj_from_dict(data_info, datasets) File "/home/cuihuili/miniconda3/envs/multiperson1/lib/python3.7/site-packages/mmcv-0.2.10-py3.7-linux-x86_64.egg/mmcv/runner/utils.py", line 81, in obj_from_dict return obj_type(**args) File "/home/cuihuili/coherent_reconstruction/multiperson-master/mmdetection/mmdet/datasets/h36m.py", line 107, in init self.img_infos = self.load_annotations(ann_file) File "/home/cuihuili/coherent_reconstruction/multiperson-master/mmdetection/mmdet/datasets/h36m.py", line 239, in load_annotations with open(ann_file, 'rb') as f: FileNotFoundError: [Errno 2] No such file or directory: 'data/h36m/extras/rcnn/h36m_train.pkl' Q1,why missing keys in source state_dict? Q2,when evaluation Panoptic, also need h36m dataset?