xingyizhou / pytorch-pose-hg-3d

PyTorch implementation for 3D human pose estimation
GNU General Public License v3.0
615 stars 141 forks source link

Error Benchmark Testing #56

Open NguyenDangBinh opened 5 years ago

NguyenDangBinh commented 5 years ago

dear, I use CPU, not GPU, not CUDA. I use Anaconda, Spider to build main.py. I had a error: I met the error like that image Can you help me?

xingyizhou commented 5 years ago

Hi, CPU training is not supported in this project.

NguyenDangBinh commented 5 years ago

Dear, I had GPU and I had the error below (tensorflow-gpu) D:\Binh\pytorch-pose-hg-3d-master\pytorch-pose-hg-3d-master>python src/main.py --exp_id test --task human3d --dataset fusion_3d --load_model models/fusion_3d_var.pth --test --full_test USE_TENSORBOARD: False heads {'hm': 16, 'depth': 16} 'cp' 不是內部或外部命令、可執行的程式或批次檔。 => using msra resnet 'msra_50' => loading pretrained model https://download.pytorch.org/models/resnet50-19c8e357.pth loaded models/fusion_3d_var.pth, epoch 10 ==> initializing 3D val data. Loaded 3D val 19312 samples

Images2D 0, #Images3D 19312

USE_TENSORBOARD: False Traceback (most recent call last): File "src/main.py", line 103, in main(opt) File "src/main.py", line 59, in main log_dict_train, preds = val(0, opt, val_loader, model) File "D:\Binh\pytorch-pose-hg-3d-master\pytorch-pose-hg-3d-master\src\lib\train_3d.py", line 147, in val_3d return step('val', epoch, opt, val_loader, model) File "D:\Binh\pytorch-pose-hg-3d-master\pytorch-pose-hg-3d-master\src\lib\train_3d.py", line 38, in step for i, batch in enumerate(data_loader): File "C:\Users\user\Anaconda3\envs\tensorflow-gpu\lib\site-packages\torch\utils\data\dataloader.py", line 637, in next return self._process_next_batch(batch) File "C:\Users\user\Anaconda3\envs\tensorflow-gpu\lib\site-packages\torch\utils\data\dataloader.py", line 658, in _process_next_batch raise batch.exc_type(batch.exc_msg) AttributeError: Traceback (most recent call last): File "C:\Users\user\Anaconda3\envs\tensorflow-gpu\lib\site-packages\torch\utils\data\dataloader.py", line 138, in _worker_loop samples = collate_fn([dataset[i] for i in batch_indices]) File "C:\Users\user\Anaconda3\envs\tensorflow-gpu\lib\site-packages\torch\utils\data\dataloader.py", line 138, in samples = collate_fn([dataset[i] for i in batch_indices]) File "D:\Binh\pytorch-pose-hg-3d-master\pytorch-pose-hg-3d-master\src\lib\datasets\fusion_3d.py", line 34, in getitem return self.dataset3D[index] File "D:\Binh\pytorch-pose-hg-3d-master\pytorch-pose-hg-3d-master\src\lib\datasets\h36m.py", line 96, in getitem s = min(s, max(img.shape[0], img.shape[1])) * 1.0 AttributeError: 'NoneType' object has no attribute 'shape'

xingyizhou commented 5 years ago

Hi, it seems that you haven't set up the image path correctly. You can add print('path') after https://github.com/xingyizhou/pytorch-pose-hg-3d/blob/master/src/lib/datasets/h36m.py#L55 to see if the path exists on your computer.

NguyenDangBinh commented 5 years ago

dear, I fix the image path but the same error. image

I download the folder like that image