how to use load_mat_into_csv_xml.py?

[TahereSaidi]

Hi, I have used this dataset and also used the repository provided by you to generate frames. But I divided the videos differently. That is, I first divided the videos into three training, evaluation and test files. As a result, there is no video with some frames in the training data and some frames in the test data.. I also put the labels of each of these three categories in separate files called train labels, evaluation labels and test labels. Now I want to convert the labels from .mat format to another format using load_mat_into_csv_xml.py file. I downloaded the load_mat_into_csv_xml.py file and used it like this:!python -i "the address of training labels" -o"the address of output", -f"the address of folder that has train and test frames" but it gives an errore: File "/Users/goly/miniconda3/lib/python3.11/site-packages/scipy/io/matlab/_mio4.py", line 179, in read_sub_array
raise ValueError("Not enough bytes to read matrix '%s'; is this "
ValueError: Not enough bytes to read matrix 'b'\x10\x0b\x
what should i do?

[tilemmpon]

I apologize for the late response. Could you please let me know if you still need assistance?

[TahereSaidi]

no thank you. solved.

[lmaple24327]

Hello, I would like to inquire about the file objects_onshore.txt, which is the tag file of the target in the video dataset On shore.
I have some problems when using this tag file, saying that it is not normalized.
The problem description looks like this:
train WARNING ⚠️ C:\Users\admin\Desktop\ultralytics-main\datasets\MVI_1622_VIS-dataset\images\train\MVI_1622_VIS240.jpg: ignoring corrupt image/label: non-normalized or out of bounds coordinates [249.81 98.713 495.62 195.43 249.81 33.699 495.62 65.398 235.28 23.324 466.57 44.648.
Do you know how to solve it? Looking forward to your reply, thank you very much.