DOTA problems

[HOHKUIK0217]

Hello, after processing the DIOR dataset, I moved on to the DOTA 1.0 dataset. Following test.dior.py, I generated JSON files for train and val sets, then proceeded to run dota_img_to_caption.py. However, I encountered an error:

bash
Traceback (most recent call last):
File "/home/xxx/RSMamba/dota_img_to_caption.py", line 234, in
i_p = position.index(obj["pos"])
^^^^^^^^^^^^^^^^^^^^^^^^^^
ValueError: tuple.index(x): x not in tuple
I'm wondering if this is because DOTA dataset images weren't processed to 512×512? I'm a bit confused about the cause.

[AZZMM]

We RSMamba to classify image scene to different categories, which are different between DIOR and DOTA dataset. For DOTA dataset, we implement test_dota.py in data_tools/caption/RSMamba folder. Using test_DIOR.py will cause conflicts. Hope this can help you.