open-mmlab/mmrotate

Issues

• [Bug] if device.type == "cpu":

  #978 opened 5 months ago by ning-feng
  7

• [Bug] ConvertMask2BoxType function has the wrong order of height and width for PolygonMasks

  #1038 opened 11 days ago by niecongchong
  0

• [Bug] The reproduced mAP is much less than the official result(RTMDet, O-RCNN, RetinaNet)/复现mAP离官方宣称结果差别较大

  #1036 opened 19 days ago by Terry-Zheng
  3

• [Need Help !!! ] how to get/debug loss with mmrotate

  #1035 opened 20 days ago by CrazyBrick
  4

• [Bug] 在验证时想要用预测值和真实值的水平框计算iou应该怎么做？

  #1037 opened 18 days ago by XDEDLS
  6

• 如何在二阶段算法中使用CSL

  #1034 opened 21 days ago by CS-Guo
  0

• [Bug] 执行projects目录下的RR360项目报unexpected keyword argument 'img_shape错误

  #1033 opened 21 days ago by yvanwy
  0

• [Bug] Can't train model with the existing configs and DOTA v1.0 in the Docker

  #1031 opened a month ago by sjhong310
  0

• [Bug] fps --task inference

  #1011 opened 3 months ago by crml233
  2

• [Docs] 太不严谨了！！错误百出！

  #1027 opened a month ago by willwang-cv
  1

• [Docs] 最新文档 部分内容 编写错误！

  #1028 opened a month ago by willwang-cv
  0

• [Docs] 什么鬼文档？各种版本根本对不上！

  #1026 opened a month ago by willwang-cv
  0

• [Bug] mim install mmdet\<3.0.0 系统找不到指定的文件。

  #1025 opened a month ago by willwang-cv
  0

• TypeError: list indices must be integers or slices, not tuple

  #1020 opened 2 months ago by tianbo321
  1

• how to add negative sample

  #1023 opened 2 months ago by nanhui69
  0

• HRSC2016舰船数据集的数量以及标注有问题

  #977 opened 6 months ago by 15689933561
  1

• When training the r3det on SSDD, the precision is too low.

  #984 opened 5 months ago by xiaxiangyu
  3

• [Help]

  #1022 opened 2 months ago by IAMLEIzZ
  0

• STD模型 训练HRSCD数据集报错：HRSCDataset: [Errno 2] No such file or directory: 'data/HRSC2016/FullDataSet/AllImages/data/HRSC2016/FullDataSet/Annotations/100000001.xml'

  #1004 opened 3 months ago by Joey-He
  3

• ssdd

  #1021 opened 2 months ago by 0Freeebaby
  1

• [Bug] CPU-ONLY TORCH, NO CUDA, Running Example From Getting Started without CUDA

  #1009 opened 2 months ago by DerKevinRiehl
  3

• [Bug] With MMrotate dev1.x branch, train rotated faster RCNN on DOTA loss=nan

  #988 opened 4 months ago by harmlessSR
  1

• mmrotate\models\losses\kf_iou_loss.py", line 21, in xy_wh_r_2_xy_sigma _shape = xywhr.shape AttributeError: 'NoneType' object has no attribute 'shape'[Bug]

  #1019 opened 2 months ago by xxxyyynnn
  0

• [Bug] RotatedBoxes.Flip_() results may not match angle_version

  #1018 opened 2 months ago by Another-0
  0

• 模型onnx转换问题

  #1017 opened 2 months ago by nanhui69
  0

• Fine-tune RTMDET On Custom Dataset!!

  #1016 opened 2 months ago by KaranBhuva22
  0

• TypeError: S2ANet: ConvNeXt: __init__() got an unexpected keyword argument 'pretrained'

  #1015 opened 2 months ago by xiaxiangyu
  1

• [Docs] We've encounterd an issue that cant link to evalution server

  #1014 opened 3 months ago by Aocide119
  2

• [Bug] DOTA Evaluation Results Feedback of Task1 is obviously lower than the official result

  #1013 opened 3 months ago by jimuIee
  4

• [Bug] H2rbox DDP training ERROR, when backbone changed to SwinTransformer

  #1012 opened 3 months ago by SyyydShen
  1

• Issue in loading pretrained weights while fine-tuning!!

  #1007 opened 3 months ago by KaranBhuva22
  1

• [Docs] how to debug backbones (like resnet or swin) or necks(like FPN) of mmdet in mmrotate projects?

  #1010 opened 3 months ago by James-zcx00
  0

• [Docs] Installation Instruction Errorneous

  #1008 opened 3 months ago by DerKevinRiehl
  0

• [Bug] 训练和测试rotated_reppoints模型的时候会出现RuntimeError: CUDA error: an illegal memory access was encountered错误

  #991 opened 4 months ago by luoz66
  1

• How can i use nms_rotated with class_agnostic=True?

  #1006 opened 3 months ago by gjustin40
  0

• [Bug] 想要将RTMDet_MMRotate的Backbone更换为Swin-Base-transformer

  #1005 opened 3 months ago by Wangfeng1971
  0

• ReduceOnPlateauLR incompatible with rtmdet?

  #1002 opened 3 months ago by tikitong
  0

• [Bug] The hrsid dataset was downloaded from the baidu web disk, but a message was displayed indicating that hrsid.z01 was missing during decompression, and the data set could not be decompressed.

  #1001 opened 3 months ago by smilyhd
  0

• [Bug] How to infer the trained model on the DOTA-1.0 dataset using the MMrotate dev1.x branch, and generate a compressed file for online submission?

  #999 opened 3 months ago by fadaishaitaiyang
  2

• [Bug] 是否支持m1 苹果芯片？

  #998 opened 4 months ago by zsk1993
  0

• Not able to analyze logs that are generated during training as provided in the documentation

  #982 opened 5 months ago by Manjuphoenix
  1

• hh2rbox

  #997 opened 4 months ago by yechao-prog
  1

• [Bug] a problem while running tools/analysis_tools/get_flops.py to get the flops of ReDet detector

  #996 opened 4 months ago by zcj234
  0

• 启动训练报错 ModuleNotFoundError: No module named 'mmdet'[Bug]

  #994 opened 4 months ago by gstariarch
  0

• img_split produces empty annotations (empty txt files) for test dataset

  #989 opened 4 months ago by arielsolomon
  0

• Is there anybody out there? - please communicate on the state of the MMRotate project.

  #987 opened 4 months ago by timresink
  0

• How to load the weights trained with mmrotate0.32 into mmrotate1.0.0rc1 for testing？

  #983 opened 5 months ago by hh123445
  0

• [Bug] can't mim install mmdet successfully

  #979 opened 5 months ago by B1029038
  0

• 问一下hrsc2016数据集的最高指标现在多少，有没有比STD高的，再问下mmrotate中map是map07还是map12啊

  #976 opened 6 months ago by 15689933561
  0

• [Bug] docker build success but run failed (segmentfalt)

  #975 opened 6 months ago by xylcbd
  0