search

aim-uofa / Poseur

[ECCV 2022] The official repo for the paper "Poseur: Direct Human Pose Regression with Transformers".
Other
179 stars 13 forks source link

keypoint_head #9

Closed flomok closed 1 year ago

flomok commented 1 year ago

hello,I haven't changed your code,but there is a problem: " File "E:/python_console/pose_2/Poseur-main1/tools/train.py", line 202, in main() File "E:/python_console/pose_2/Poseur-main1/tools/train.py", line 198, in main meta=meta) File "E:\python_console\pose_2\Poseur-main\mmpose\apis\train.py", line 212, in train_model runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "E:\ana\envs\ptorch\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 136, in run epoch_runner(data_loaders[i], kwargs) File "E:\ana\envs\ptorch\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 53, in train self.run_iter(data_batch, train_mode=True, kwargs) File "E:\ana\envs\ptorch\lib\site-packages\mmcv\runner\epoch_based_runner.py", line 32, in run_iter kwargs) File "E:\ana\envs\ptorch\lib\site-packages\mmcv\parallel\data_parallel.py", line 77, in train_step return self.module.train_step(inputs[0], kwargs[0]) File "E:\python_console\pose_2\Poseur-main\mmpose\models\detectors\base.py", line 95, in train_step losses = self.forward(data_batch) File "E:\ana\envs\ptorch\lib\site-packages\mmcv\runner\fp16_utils.py", line 146, in new_func output = old_func(new_args, new_kwargs) File "E:\python_console\pose_2\Poseur-main\mmpose\models\detectors\poseur.py", line 92, in forward *kwargs) File "E:\python_console\pose_2\Poseur-main\mmpose\models\detectors\poseur.py", line 115, in forward_train output = self.keypoint_head(output, img_metas) File "E:\ana\envs\ptorch\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl return forward_call(input, **kwargs) TypeError: forward() takes 2 positional arguments but 3 were given" what is the keypoint_head's input?

YongtaoGe commented 1 year ago

Hi, @flomok, the input of keypoint head is the multi-level features from the neck. You can pull the code from the main branch and try it again. We have fixed the problem.