fasterrcnn-pytorch-training-pipeline icon indicating copy to clipboard operation
fasterrcnn-pytorch-training-pipeline copied to clipboard

COCO detetction Dataset

Open anirudh6415 opened this issue 1 year ago • 2 comments

Hi, I have been trying to use https://pytorch.org/vision/stable/_modules/torchvision/datasets/coco.html to load the data and use it for training but model is not learning anything how do i work with it. Can you help me with it.

Thank you

anirudh6415 avatar Feb 12 '24 19:02 anirudh6415

Hi. The COCO dataset YAML file in the data_configs directory expects 81 classes instead of 91 classes. You can find the correct dataset here on Kaggle. If you are facing a particular error, please let me know.

sovit-123 avatar Feb 13 '24 00:02 sovit-123

Hi Sovit, I wanted to know, If i can use

from torchvision.datasets import CocoDetection ,wrap_dataset_for_transforms_v2
dataset = CocoDetection(root_dir, annotation_file,transforms=transforms)

but model is not learning anything how do i work with it. Can you help me with it.

anirudh6415 avatar Feb 13 '24 01:02 anirudh6415