Parse the pkl of results Now, I have trained the model and get the results.pkl.

[houliping]

How to parse the results.pkl to the format needed for DOTA evaluation (Class AP and mAP )? Looking forward for your reply, thanks

[LiWentomng]

First, you can refer to this script parse_pkl_for_evaluation.py to parse the pkl file to get the standard text format.

Then, you can run the code to get the mAP and class APs, and run the code to get mAOE and class AOEs for evaluation of val dataset(train set for training) . And you need submit to the DOTA evaluation sever to get the results for evaluation of test dataset(train and val set for training).

Finally, you can refer to the visual code to show the learning points and oriented boxes.

[yzk-lab]

How to parse the results.pkl to the format needed for DOTA evaluation (Class AP and mAP )? Looking forward for your reply, thanks.

你好，请问你有解决这个pkl的parse的问题吗？parse之后是裁剪的图片的形式，没有对图片进行合并。

[yzk-lab]

我在这个parse_pkl里好像没有看到result_merge_multi_process进行合并的操作，您好，请问这个合并是要怎么进行呢？现在我得到了训练模型，进行了pkl的parse,得到一个result_raw文件夹但是提交到DOTA官网出错，应该是没有合并的缘故，你有什么好的解决办法吗？谢谢回复！

我之前尝试的S2anet是直接test可以生成result_raw和result_merge，现在想在您的code里也得到result_merge，但是好像出错了，希望得到您的回复，感谢！

[LiWentomng]

是的， parse_pkl_for_evaluation.py生成的是results_raw, 对于dota results_raw的合并，请参考该脚本dota_merge_results.py 。

[yzk-lab]

是的， parse_pkl_for_evaluation.py生成的是results_raw, 对于dota results_raw的合并，请参考该脚本dota_merge_results.py

谢谢！我试一试！

[H1d30nbu3h]

How to parse the results.pkl to the format needed for DOTA evaluation (Class AP and mAP )? Looking forward for your reply, thanks

您好，我再用test生成pkl文件的时候出现了一个bug ，for i, data in enumerate(data_loader): #，报了RecursionError: maximum recursion depth exceeded while calling a Python object溢出的错误，请问您能帮我解答一下吗