abdullahtarek / tennis_analysis Goto Github PK

Hello, I am a student. Your method is very interesting. Can you provide me with the model folder for testing? Thanks

There are 2 typos in https://github.com/abdullahtarek/tennis_analysis/blob/main/training/tennis_court_keypoints_training.ipynb

1. kps = np.array(items['kps']).flatten() should be kps = np.array(item['kps']).flatten()
2. devic = torch.device("cuda" if torch.cuda.is_available() else "cpu") should be device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

Hi, Love the YouTube Channel. When Creating the Torch Data Set, there seems to be a bug when creating the np.array for the Keys with kps. In particular kps = np.array(items['kps']).flatten() in the getitem(self, idx) method. Thanks!

cause the basketball court is very wide, the camera cannot include all keypoints in one image, how can I do, any help

hi, please help, I encountered these errors even after changing the dataset details:
raise RuntimeError(emojis(f"Dataset '{clean_url(self.args.data)}' error ❌ {e}")) from e
RuntimeError: Dataset 'C://Users/My name is Krish/Downloads/TENNIS_ANALYSIS/training/Tennis-Ball-Detection-5/data.yaml' error
Dataset 'C://Users/My name is Krish/Downloads/TENNIS_ANALYSIS/training/Tennis-Ball-Detection-5/data.yaml' images not found , missing path 'C:\Users\My name is Krish\Downloads\TENNIS_ANALYSIS\training\Tennis-Ball-Detection-5\valid\images'
Note dataset download directory is 'C:\Users\My name is Krish\Downloads\TENNIS_ANALYSIS\datasets'. You can update this in 'C:\Users\My name is Krish\AppData\Roaming\Ultralytics\settings.yaml'

THE ABOVE CODE IS NOT BEING EXECUTED, i am a student and is highly relaying on works like yours to self-learn ML . could you please upload the "yolo5_best.pt" and "yolo5_last.pt" files in model folder in git hub. it would be so helpful as i am currently stuck in that step.