search

microsoft / human-pose-estimation.pytorch

The project is an official implement of our ECCV2018 paper "Simple Baselines for Human Pose Estimation and Tracking(https://arxiv.org/abs/1804.06208)"
MIT License
2.93k stars 601 forks source link

I have an error when I load image file in data/coco/images/val2017 #175

Open sun-tao opened 2 years ago

sun-tao commented 2 years ago

when I use !python3 pose_estimation/valid.py \ --cfg experiments/coco/resnet50/256x192_d256x3_adam_lr1e-3.yaml \ --flip-test \ --model-file models/pytorch/pose_coco/pose_resnet_50_256x192.pth.tar\ --workers 2 to valid the pretrained model,the following error happens!

=> classes: ['background', 'person'] => num_images: 5000 => load 6352 samples => fail to read data/coco/images/val2017/000000397133.jpg Traceback (most recent call last): File "pose_estimation/valid.py", line 168, in => fail to read data/coco/images/val2017/000000476258.jpg main() File "pose_estimation/valid.py", line 164, in main final_output_dir, tb_log_dir) File "/content/drive/MyDrive/simplepose /human-pose-estimation.pytorch/pose_estimation/../lib/core/function.py", line 108, in validate for i, (input, target, target_weight, meta) in enumerate(val_loader): File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py", line 521, in next data = self._next_data() File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py", line 1203, in _next_data return self._process_data(data) File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py", line 1229, in _process_data data.reraise() File "/usr/local/lib/python3.7/dist-packages/torch/_utils.py", line 434, in reraise raise exception ValueError: Caught ValueError in DataLoader worker process 0. Original Traceback (most recent call last): File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/_utils/worker.py", line 287, in _worker_loop data = fetcher.fetch(index) File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/_utils/fetch.py", line 49, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/_utils/fetch.py", line 49, in data = [self.dataset[idx] for idx in possibly_batched_index] File "/content/drive/MyDrive/simplepose /human-pose-estimation.pytorch/pose_estimation/../lib/dataset/JointsDataset.py", line 80, in getitem raise ValueError('Fail to read {}'.format(image_file)) ValueError: Fail to read data/coco/images/val2017/000000397133.jpg

Dipankar1997161 commented 2 years ago

This error occurs when the code is unable to locate your image file.

I encountered this when using my own dataset. The best way to sort this out is to keep that particular image outside the Val/Train folder and run the code again. For example: for Valid images. Instead of placing them inside /data/images/valid2017. Just keep it /data/images and then try. Try with 1 image and see if the same error occurs for the next image of the valid. Then you would understand.

It worked for me!

zorica-zdravkovska commented 11 months ago

@Dipankar1997161 Hello can you please explain how did you use the model for your own dataset?