-
run RRPN_demo.py occurs
from demo.predictor import ICDARDemo, RRPNDemo
: cannot import name 'ICDARDemo'
I find three classes in demo.predictor.I can import COCODemo,but can't import ICDARDemo,…
-
最近在研究mask R-cnn，对您发的论文特别感兴趣，请问有没有办法将您的RRPN移植到mask R-cnn中啊
-
Hello. When I use "e2e_rrpn_R_50_C4_1x_ICDAR13_15_trial.yaml" to train IC15 data , I found that loss_objectness and loss_rpn_box_reg is always nan. I have tried to reduce base_lr to 0.0001 but it didn…
-
When I predict a text area, rotate image and crop it, I often find that if a text area`s angle is 225°, the rrpn predict 45°. After I input the crop image croped by 45°, the crnn result goes wrong. So…
-
## ❓ Questions and Help
Hey，thank you for sharing these codes. What puzzles me is why the label of the bbox generated in the convert_to_coco.py file is a horizontal rectangle instead of a oriented re…
-
Hi
I am unable to find a source to obtain './data_cache/alphabet_IC13_IC15_Syn800K_pro.txt'.
I understand that training the network generates the file at ./data_cache folder as mentioned in #3 but …
-
FPN的结构，每层特征图anchor size的大小从下往上应该是３２，６４，１２８，２５６，５１２．对应的anchor stride是４，８，１６，３２，６４．
配置文件中的设置是： ANCHOR_STRIDE: (4, 8, 16, 32, 64)，ANCHOR_SIZES: (2, 4, 8, 16, 32)
配置文件中的解释是ANCHOR_SIZES里面的值是１６的倍数，但是运行的代码…
-
Hi,
I encountered the problem of CUDA_ERROR_ILLEGAL_ADDRESS error during training when the objects are densely located, i.e. when there are many(roughly >100 or 150) ships in one 600*1000 image. Wh…
-
我使用 `./experiment/scripts/faster_rcnn_end2end.sh 1 VGG16 rrpn` 来训练模型。训练阶段没有问题，成功得到了最终的模型，但是测试阶段出现了错误，是 `test_net.py` 中调用 `get_imdb` 函数时出的错， 提示 `KeyError: 'Unknown dataset: MSRA_TEST'，请问我应该如何解决这个错误？
-
hello,I use my own training data ,when i run the script ./experiments/scripts/faster_rcnn_end2end.sh 0 VGG16 rrpn.i got the error:
./experiments/scripts/faster_rcnn_end2end.sh: line 78: 38172 Float…