AerialDetection
AerialDetection copied to clipboard
dingjiansw101
Ding，Thx for your contribution. and I have one issue about you model. That is when I use it to train HRSC2016 Datasets. what should I focus on ? and how...
First of all, I thank you for creating such a amazing toolkit. My data set consists of aerial images which is different from DOTA. I convert my dataset to DOTA...
多尺度 用了哪些尺度 gap512是指 prepare_dota1.py和prepare_dota1_aug.py里gap的参数均为512吗
When i was training, the memory has ran out. My env: > sys.platform: linux > Python: 3.7.4 (default, Aug 13 2019, 20:35:49) [GCC 7.3.0] > CUDA available: True > CUDA_HOME:...
Hey, Thank you for the repo, I am using this repo to learn and train my own model on rotated images dataset. I trained the model on my images successfully...
running build_ext building 'poly_nms_cuda' extension creating /maxiangjun/AerialDetection/mmdet/ops/poly_nms/build creating /maxiangjun/AerialDetection/mmdet/ops/poly_nms/build/temp.linux-x86_64-3.8 creating /maxiangjun/AerialDetection/mmdet/ops/poly_nms/build/temp.linux-x86_64-3.8/src Emitting ninja build file /maxiangjun/AerialDetection/mmdet/ops/poly_nms/build/temp.linux-x86_64-3.8/build.ninja... Compiling objects... Allowing ninja to set a default number of workers... (overridable by setting...
loading annotations into memory... Done (t=0.00s) creating index... index created! 2022-11-21 16:08:57,399 - INFO - Start running, host: pdrl@pdrl, work_dir: /home/pdrl/aerialdetection/AerialDetection/work_dirs/faster_rcnn_RoITrans_r50_fpn_1x_dota 2022-11-21 16:08:57,399 - INFO - workflow: [('train', 1)], max:...
My envirements are: Ubuntu18.0 GCC5.5.0 CUDA10.0 Python3.7 Pytorch==1.2.0 Torchvision==0.4 mmcv==0.4.3 I have Installed as the offical docs when I excute python demo_large_image.py it stuck at 0% for more than 10...
when i run python demo.large.py i get this error. raise IOError('{} is not a checkpoint file'.format(filename)) OSError: work_dirs/faster_rcnn_RoITrans_r50_fpn_1x_dota/epoch_12.pth is not a checkpoint file i suppose this is where u put...
# Welcome update to OpenMMLab 2.0 I am Vansin, the technical operator of OpenMMLab. In September of last year, we announced the release of OpenMMLab 2.0 at the World Artificial...
