Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

DOTA evaluation命令根本无法执行 #16

Open
doctorcoal opened this issue Jun 19, 2024 · 1 comment
Open

DOTA evaluation命令根本无法执行 #16

doctorcoal opened this issue Jun 19, 2024 · 1 comment

Comments

@doctorcoal
Copy link

使用所有你在README.md中提供的信息也根本不可能在DOTA上正常运行
即使我准备好了官方格式的DOTA,它会报错FileNotFoundError: [Errno 2] No such file or directory: '/data/DOTA/val/valset.txt'
全网没有任何人对这个文件提供一个官方的说法
即使我通过手写来实现这个文件,它依然会提示我FileNotFoundError: [Errno 2] No such file or directory: './work_dirs/Task1_rotated_faster_rcnn_r50_fpn_1x_dota_le90_epoch_12/Task1_plane.txt'
即使我按照DOTA官网的格式准备好了每个文件,它依然会报错:
classname: plane
Traceback (most recent call last):
File "./dota_evaluation_task1.py", line 324, in
main(opt.mergedir, opt.imagesetdir, opt.use_07_metric)
File "./dota_evaluation_task1.py", line 300, in main
use_07_metric=use_07_metric)
File "./dota_evaluation_task1.py", line 179, in voc_eval
BB = BB[sorted_ind, :]
IndexError: too many indices for array: array is 1-dimensional, but 2 were indexed
非常糟糕的开源
任何一个看到我这个issue的人,我建议你立刻打消使用这个开源来在DOTA上跑的计划,避免浪费大把的时间

@yuhongtian17
Copy link
Owner

yuhongtian17 commented Jun 19, 2024

你的代码能力很松弛,但是你的装逼欲很好地弥补了这一点。如果你看懂了DOTA devkitdota_evaluation_task1.py代码你就不会问出这样的问题,你就应该知道,valset.txt只是一个列举出val集图片名的文本文件。大家都惯例这么使用的,都能理解出来和写出来,这到底是谁的问题呢?你要是写不出来的话我这就帮你找一个
还有,Task1_plane.txt这些文件就根本不需要手动去准备的。我在example usage里告诉你了,先用

CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 ./tools/dist_train.sh ./configs/rotated_faster_rcnn/rotated_faster_rcnn_r50_fpn_1x_dota_le90.py 8

训一遍,再用

CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 ./tools/dist_test.sh ./configs/rotated_faster_rcnn/rotated_faster_rcnn_r50_fpn_1x_dota_le90.py ./work_dirs/rotated_faster_rcnn_r50_fpn_1x_dota_le90/epoch_12.pth 8 --format-only --eval-options submission_dir="./work_dirs/Task1_rotated_faster_rcnn_r50_fpn_1x_dota_le90_epoch_12/"

生成出一个存储结果的文件夹,这个文件夹里就已经做好了Task1_plane.txt、Task1_ship.txt这些文件。最后用

python "../DOTA_devkit/dota_evaluation_task1.py" --mergedir "./work_dirs/Task1_rotated_faster_rcnn_r50_fpn_1x_dota_le90_epoch_12/" --imagesetdir "./data/DOTA/val/" --use_07_metric True

去合并到原图上进行评测。我劝你务实一点,先把DOTA数据集目标检测的流程搞搞懂,MMRotate写的挺好的,你去把它扔了干什么?自己手写个空的Task1_plane.txt文本文件结果报错,你倒告诉我,怎么回答你这个issue呢?——————

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants