The results of retinanet-obb cannot be reproduced.

[zytx121]

The mAP we got only 66.95, not 69 as you said.

mAP: 0.6695016721761877 ap of each class: plane:0.8933495268168008, baseball-diamond:0.7374626613068532, bridge:0.4122850723332845, ground-track-field:0.6670794506088387, small-vehicle:0.760291145560589, large-vehicle:0.6047573582066816, ship:0.7630074552316077, tennis-court:0.9088855421686749, basketball-court:0.7977999161008033, storage-tank:0.8171024978735246, soccer-ball-field:0.5278917624380285, roundabout:0.635699247940088, harbor:0.514493570948177, swimming-pool:0.6264466877239606, helicopter:0.3759731873849022

I don't know what went wrong?

Here is our cfg:

2021-10-21 15:01:46,198 - mmdet - INFO - Environment info:

sys.platform: linux Python: 3.7.7 (default, Mar 23 2020, 22:36:06) [GCC 7.3.0] CUDA available: True CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 10.1, V10.1.243 GPU 0: GeForce RTX 2080 Ti GCC: gcc (Ubuntu 7.4.0-1ubuntu1~18.04.1) 7.4.0 PyTorch: 1.7.1 PyTorch compiling details: PyTorch built with:

• GCC 7.3
• C++ Version: 201402
• Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v1.6.0 (Git Hash 5ef631a030a6f73131c77892041042805a06064f)
• OpenMP 201511 (a.k.a. OpenMP 4.5)
• NNPACK is enabled
• CPU capability usage: AVX2
• CUDA Runtime 10.2
• NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75
• CuDNN 7.6.5
• Magma 2.5.2
• Build settings: BLAS=MKL, BUILD_TYPE=Release, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DUSE_VULKAN_WRAPPER -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, USE_CUDA=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON,

TorchVision: 0.8.2 OpenCV: 4.2.0 MMCV: 0.6.2 MMDetection: 2.2.0+unknown MMDetection Compiler: GCC 7.4 MMDetection CUDA Compiler: 10.1

2021-10-21 15:01:46,199 - mmdet - INFO - Distributed training: True 2021-10-21 15:01:46,373 - mmdet - INFO - Config: dataset_type = 'DOTADataset' data_root = '/data/zy/datasets/split_ss_dota1_0/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict( type='LoadOBBAnnotations', with_bbox=True, with_label=True, with_poly_as_mask=True), dict(type='LoadDOTASpecialInfo'), dict(type='Resize', img_scale=(1024, 1024), keep_ratio=True), dict(type='OBBRandomFlip', h_flip_ratio=0.5, v_flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict( type='RandomOBBRotate', rotate_after_flip=True, angles=(0, 0), vert_rate=0.5, vert_cls=['roundabout', 'storage-tank']), dict(type='Pad', size_divisor=32), dict(type='DOTASpecialIgnore', ignore_size=2), dict(type='FliterEmpty'), dict(type='Mask2OBB', obb_type='obb'), dict(type='OBBDefaultFormatBundle'), dict( type='OBBCollect', keys=['img', 'gt_bboxes', 'gt_obboxes', 'gt_labels']) ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipRotateAug', img_scale=[(1024, 1024)], h_flip=False, v_flip=False, rotate=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='OBBRandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='RandomOBBRotate', rotate_after_flip=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='OBBCollect', keys=['img']) ]) ] data = dict( samples_per_gpu=2, workers_per_gpu=4, train=dict( type='DOTADataset', task='Task1', ann_file='/data/zy/datasets/split_ss_dota1_0/trainval/annfiles/', img_prefix='/data/zy/datasets/split_ss_dota1_0/trainval/images/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='LoadOBBAnnotations', with_bbox=True, with_label=True, with_poly_as_mask=True), dict(type='LoadDOTASpecialInfo'), dict(type='Resize', img_scale=(1024, 1024), keep_ratio=True), dict(type='OBBRandomFlip', h_flip_ratio=0.5, v_flip_ratio=0.5), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict( type='RandomOBBRotate', rotate_after_flip=True, angles=(0, 0), vert_rate=0.5, vert_cls=['roundabout', 'storage-tank']), dict(type='Pad', size_divisor=32), dict(type='DOTASpecialIgnore', ignore_size=2), dict(type='FliterEmpty'), dict(type='Mask2OBB', obb_type='obb'), dict(type='OBBDefaultFormatBundle'), dict( type='OBBCollect', keys=['img', 'gt_bboxes', 'gt_obboxes', 'gt_labels']) ]), test=dict( type='DOTADataset', task='Task1', ann_file='/data/zy/datasets/split_ss_dota1_0/test/annfiles/', img_prefix='/data/zy/datasets/split_ss_dota1_0/test/images/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipRotateAug', img_scale=[(1024, 1024)], h_flip=False, v_flip=False, rotate=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='OBBRandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='RandomOBBRotate', rotate_after_flip=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='OBBCollect', keys=['img']) ]) ])) evaluation = None optimizer = dict(type='SGD', lr=0.0025, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2)) lr_config = dict( policy='step', warmup='linear', warmup_iters=500, warmup_ratio=0.001, step=[8, 11]) total_epochs = 12 checkpoint_config = dict(interval=1) log_config = dict(interval=50, hooks=[dict(type='TextLoggerHook')]) dist_params = dict(backend='nccl') log_level = 'INFO' load_from = None resume_from = None workflow = [('train', 1)] model = dict( type='RetinaNetOBB', pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, norm_cfg=dict(type='BN', requires_grad=True), norm_eval=True, style='pytorch'), neck=dict( type='FPN', in_channels=[256, 512, 1024, 2048], out_channels=256, start_level=1, add_extra_convs='on_input', num_outs=5), bbox_head=dict( type='OBBRetinaHead', num_classes=15, in_channels=256, stacked_convs=4, feat_channels=256, anchor_generator=dict( type='Theta0AnchorGenerator', octave_base_scale=4, scales_per_octave=3, ratios=[0.5, 1.0, 2.0], strides=[8, 16, 32, 64, 128]), bbox_coder=dict( type='OBB2OBBDeltaXYWHTCoder', target_means=[0.0, 0.0, 0.0, 0.0, 0.0], target_stds=[1.0, 1.0, 1.0, 1.0, 1.0]), loss_cls=dict( type='FocalLoss', use_sigmoid=True, gamma=2.0, alpha=0.25, loss_weight=1.0), loss_bbox=dict(type='L1Loss', loss_weight=1.0))) train_cfg = dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.4, min_pos_iou=0, gpu_assign_thr=-1, ignore_iof_thr=-1, iou_calculator=dict(type='OBBOverlaps')), allowed_border=-1, pos_weight=-1, debug=False) test_cfg = dict( nms_pre=2000, min_bbox_size=0, score_thr=0.05, nms=dict(type='obb_nms', iou_thr=0.1), max_per_img=2000) work_dir = './work_dirs/retinanet_obb_r50_fpn_1x_dota10' gpu_ids = range(0, 1)

[jbwang1997]

It seems you use the multi-gpu training. Did you scale the learning rate based on the linear scaling rule?

[zytx121]

It seems you use the multi-gpu training. Did you scale the learning rate based on the linear scaling rule?

I did use dist_train.sh to train, but i only use 1 gpu. I think it is the same with python train.py. Did I get it wrong?

I found you have an update about opencv-5.2.1, and my opencv is 4.2.0, so i traced the code back to before this update and got 67.5 results. However it still cannot reach 69.

[jbwang1997]

I think the vision of opencv cannot influent the results. The config is the same as mine. So I think the data preparation may have some wrong.

We use train and val set, about 12800 images after splitting, to train retinanet-obb. One epoch has 6400 iterations with batch size being 2. Are we training conditions the same?

[zytx121]

why opencv cannot influent the results? I'm a little confused.

[jbwang1997]

The update for OpenCV is because definitions of obb in cv2.minAreaRect are different in opencv-5.2.1 and opencv-4.2.0.

The last commit of OBBDetection is compatible with both visions of OpenCV.

[zytx121]

The update for OpenCV is because definitions of obb in cv2.minAreaRect are different in opencv-5.2.1 and opencv-4.2.0.

The last commit of OBBDetection is compatible with both visions of OpenCV.

That's what i need !! Thx.