search

open-mmlab / mmdetection

OpenMMLab Detection Toolbox and Benchmark
https://mmdetection.readthedocs.io
Apache License 2.0
29.61k stars 9.47k forks source link

ValueError: need at least one array to concatenate #3976

Closed DorothysDad closed 3 years ago

DorothysDad commented 4 years ago

when i run python tools/train.py ./configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py

i first create VOCdevkit---VOC2007 and it includes annotations imagesets and JPEGImages, then i change faster_rcnn_r50_fpn_1x_coco.py , voc712.py, faster_rcnn_r50_fpn.py ,class_names.py and voc.py and there are mainly changes _num_classes=1 def vocclasses(): return ['crane', ] CLASSES = ('crane', )

### my voc712.py

dataset settings

dataset_type = 'VOCDataset' data_root = 'data/VOCdevkit/' 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='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1000, 600), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict(type='Normalize', img_norm_cfg), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']), ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict(type='Normalize', img_norm_cfg), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']), ]) ] data = dict( samples_per_gpu=2, workers_per_gpu=2, train=dict( type='RepeatDataset', times=3,

dataset=dict(

    #     type=dataset_type,
    #     ann_file=[
    #         data_root + 'VOC2007/ImageSets/Main/trainval.txt',
    #         data_root + 'VOC2012/ImageSets/Main/trainval.txt'
    #     ],
    #     img_prefix=[data_root + 'VOC2007/', data_root + 'VOC2012/'],
    #     pipeline=train_pipeline)),
    # 把含有VOC2012的路径去掉
    dataset=dict(
        type=dataset_type,
        ann_file=[
            data_root + 'VOC2007/ImageSets/Main/trainval.txt',
        ],
        img_prefix=[data_root + 'VOC2007/'],
        pipeline=train_pipeline)),
val=dict(
    type=dataset_type,
    ann_file=data_root + 'VOC2007/ImageSets/Main/test.txt',
    img_prefix=data_root + 'VOC2007/',
    pipeline=test_pipeline),
test=dict(
    type=dataset_type,
    ann_file=data_root + 'VOC2007/ImageSets/Main/test.txt',
    img_prefix=data_root + 'VOC2007/',
    pipeline=test_pipeline))

evaluation = dict(interval=1, metric='mAP')

### then command window gets error like that: fatal: Not a git repository (or any parent up to mount point /home) Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set). 2020-10-21 20:48:34,740 - mmdet - INFO - Environment info:

sys.platform: linux Python: 3.6.10 |Anaconda, Inc.| (default, May 8 2020, 02:54:21) [GCC 7.3.0] CUDA available: True GPU 0: GeForce GTX 1050 Ti CUDA_HOME: /usr/local/cuda-9.2 NVCC: Cuda compilation tools, release 9.2, V9.2.148 GCC: gcc (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609 PyTorch: 1.5.1+cu92 PyTorch compiling details: PyTorch built with:

TorchVision: 0.6.1+cu92 OpenCV: 4.4.0 MMCV: 1.1.5 MMCV Compiler: GCC 7.3 MMCV CUDA Compiler: 9.2 MMDetection: 2.5.0+

2020-10-21 20:48:35,042 - mmdet - INFO - Distributed training: False 2020-10-21 20:48:35,342 - mmdet - INFO - Config: model = dict( type='FasterRCNN', 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, num_outs=5), rpn_head=dict( type='RPNHead', in_channels=256, feat_channels=256, anchor_generator=dict( type='AnchorGenerator', scales=[8], ratios=[0.5, 1.0, 2.0], strides=[4, 8, 16, 32, 64]), bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[1.0, 1.0, 1.0, 1.0]), loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), loss_bbox=dict(type='L1Loss', loss_weight=1.0)), roi_head=dict( type='StandardRoIHead', bbox_roi_extractor=dict( type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0), out_channels=256, featmap_strides=[4, 8, 16, 32]), bbox_head=dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=1, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.1, 0.1, 0.2, 0.2]), reg_class_agnostic=False, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='L1Loss', loss_weight=1.0)))) train_cfg = dict( rpn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.3, min_pos_iou=0.3, match_low_quality=True, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=256, pos_fraction=0.5, neg_pos_ub=-1, add_gt_as_proposals=False), allowed_border=-1, pos_weight=-1, debug=False), rpn_proposal=dict( nms_across_levels=False, nms_pre=2000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False)) test_cfg = dict( rpn=dict( nms_across_levels=False, nms_pre=1000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100)) dataset_type = 'VOCDataset' data_root = 'data/VOCdevkit/' 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='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1000, 600), keep_ratio=True), dict(type='RandomFlip', 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='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ] data = dict( samples_per_gpu=2, workers_per_gpu=2, train=dict( type='RepeatDataset', times=3, dataset=dict( type='VOCDataset', ann_file=['data/VOCdevkit/VOC2007/ImageSets/Main/trainval.txt'], img_prefix=['data/VOCdevkit/VOC2007/'], pipeline=[ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1000, 600), keep_ratio=True), dict(type='RandomFlip', 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='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ])), val=dict( type='VOCDataset', ann_file='data/VOCdevkit/VOC2007/ImageSets/Main/test.txt', img_prefix='data/VOCdevkit/VOC2007/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ]), test=dict( type='VOCDataset', ann_file='data/VOCdevkit/VOC2007/ImageSets/Main/test.txt', img_prefix='data/VOCdevkit/VOC2007/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ])) evaluation = dict(interval=1, metric='mAP') optimizer = dict(type='SGD', lr=0.02, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=None) 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)] work_dir = './work_dirs/faster_rcnn_r50_fpn_1x_coco' gpu_ids = range(0, 1)

2020-10-21 20:48:35,651 - mmdet - INFO - load model from: torchvision://resnet50 2020-10-21 20:48:35,828 - mmdet - WARNING - The model and loaded state dict do not match exactly

unexpected key in source state_dict: fc.weight, fc.bias

fatal: Not a git repository (or any parent up to mount point /home) Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set). 2020-10-21 20:48:38,869 - mmdet - INFO - Start running, host: allen@allen, work_dir: /home/allen/mmdetection/work_dirs/faster_rcnn_r50_fpn_1x_coco 2020-10-21 20:48:38,869 - mmdet - INFO - workflow: [('train', 1)], max: 12 epochs Traceback (most recent call last): File "tools/train.py", line 178, in main() File "tools/train.py", line 174, in main meta=meta) File "/home/allen/anaconda3/lib/python3.6/site-packages/mmdet-2.5.0-py3.6.egg/mmdet/apis/train.py", line 143, in train_detector runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/home/allen/anaconda3/lib/python3.6/site-packages/mmcv/runner/epoch_based_runner.py", line 125, in run epoch_runner(data_loaders[i], **kwargs) File "/home/allen/anaconda3/lib/python3.6/site-packages/mmcv/runner/epoch_based_runner.py", line 47, in train for i, data_batch in enumerate(self.data_loader): File "/home/allen/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 279, in iter return _MultiProcessingDataLoaderIter(self) File "/home/allen/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 746, in init self._try_put_index() File "/home/allen/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 861, in _try_put_index index = self._next_index() File "/home/allen/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 339, in _next_index return next(self._sampler_iter) # may raise StopIteration File "/home/allen/anaconda3/lib/python3.6/site-packages/torch/utils/data/sampler.py", line 200, in iter for idx in self.sampler: File "/home/allen/anaconda3/lib/python3.6/site-packages/mmdet-2.5.0-py3.6.egg/mmdet/datasets/samplers/group_sampler.py", line 36, in iter indices = np.concatenate(indices) File "<__array_function__ internals>", line 6, in concatenate ValueError: need at least one array to concatenate

duongnv0499 commented 4 years ago

I had the same issue. I used my own coco format dataset and then I changed the classes in mmdet/datasets/coco.py. fortunately, it has run, but run too fast, about 10seconds/epoch. It turn out the whole training process ran within about 5mins (12 epoch). the result is extremely bad, I dont know why, please help me @RyanXLi. beside I said it run fortunately because i have changed the classes in mmdet/datasets/coco.py and in core/evaluation/class_names.py and it produce the 'need at least one array to concatenate' error. my config file : `/content/mmdetection 2020-10-21 14:56:13,079 - mmdet - INFO - Environment info:

sys.platform: linux Python: 3.6.9 (default, Jul 17 2020, 12:50:27) [GCC 8.4.0] CUDA available: True GPU 0: Tesla P100-PCIE-16GB CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 10.1, V10.1.243 GCC: gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0 PyTorch: 1.6.0+cu101 PyTorch compiling details: PyTorch built with:

TorchVision: 0.7.0+cu101 OpenCV: 4.1.2 MMCV: 1.1.5 MMCV Compiler: GCC 7.5 MMCV CUDA Compiler: 10.1 MMDetection: 2.5.0+9caa197

2020-10-21 14:56:13,421 - mmdet - INFO - Distributed training: False 2020-10-21 14:56:13,684 - mmdet - INFO - Config: model = dict( type='FasterRCNN', 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, num_outs=5), rpn_head=dict( type='RPNHead', in_channels=256, feat_channels=256, anchor_generator=dict( type='AnchorGenerator', scales=[8], ratios=[0.5, 1.0, 2.0], strides=[4, 8, 16, 32, 64]), bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[1.0, 1.0, 1.0, 1.0]), loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), loss_bbox=dict(type='L1Loss', loss_weight=1.0)), roi_head=dict( type='StandardRoIHead', bbox_roi_extractor=dict( type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0), out_channels=256, featmap_strides=[4, 8, 16, 32]), bbox_head=dict( type='Shared2FCBBoxHead', in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=5, bbox_coder=dict( type='DeltaXYWHBBoxCoder', target_means=[0.0, 0.0, 0.0, 0.0], target_stds=[0.1, 0.1, 0.2, 0.2]), reg_class_agnostic=False, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='L1Loss', loss_weight=1.0)))) train_cfg = dict( rpn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.3, min_pos_iou=0.3, match_low_quality=True, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=256, pos_fraction=0.5, neg_pos_ub=-1, add_gt_as_proposals=False), allowed_border=-1, pos_weight=-1, debug=False), rpn_proposal=dict( nms_across_levels=False, nms_pre=2000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, match_low_quality=False, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False)) test_cfg = dict( rpn=dict( nms_across_levels=False, nms_pre=1000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100)) dataset_type = 'CocoDataset' classes = ('table', 'figure', 'natural_image', 'logo', 'signature') data_root = 'data/coco/' 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='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1333, 800), keep_ratio=True), dict(type='RandomFlip', 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='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1333, 800), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ] data = dict( samples_per_gpu=2, workers_per_gpu=2, train=dict( type='CocoDataset', ann_file='data/coco/annotations/instances_train2017.json', img_prefix='data/coco/train2017/', pipeline=[ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1333, 800), keep_ratio=True), dict(type='RandomFlip', 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='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']) ]), val=dict( type='CocoDataset', ann_file='data/coco/annotations/instances_val2017.json', img_prefix='data/coco/val2017/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1333, 800), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ]), test=dict( type='CocoDataset', ann_file='data/coco/annotations/instances_val2017.json', img_prefix='data/coco/val2017/', pipeline=[ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1333, 800), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict( type='Normalize', mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']) ]) ])) evaluation = dict(interval=1, metric='bbox') optimizer = dict(type='SGD', lr=0.02, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=None) 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)] work_dir = './work_dirs/faster_rcnn_r50_fpn_1x_coco' gpu_ids = range(0, 1) `

zhang5957 commented 4 years ago

i have the same question

DorothysDad commented 4 years ago

i have the same question

do you solve it?if you do, please help me thanks

zhang5957 commented 4 years ago

May be one of your annotations file’s coordinate is null ，mmdetection doesn’t support this，each xml or json file must contain coordinate or categories. every picture must be tagged

发自我的iPhone

------------------ Original ------------------ From: Allen <notifications@github.com> Date: Mon,Nov 2,2020 8:37 PM To: open-mmlab/mmdetection <mmdetection@noreply.github.com> Cc: zhang5957 <595707874@qq.com>, Comment <comment@noreply.github.com> Subject: Re: [open-mmlab/mmdetection] ValueError: need at least one array to concatenate (#3976)

zz10001 commented 4 years ago

have you solved it? I met the same problem.

DorothysDad commented 4 years ago

have you solved it? I met the same problem.

yes, i changed my dataset and tried it then i can run it sucessfully, maybe something run in your dataset

zz10001 commented 4 years ago

yes, i changed my dataset and tried it then i can run it sucessfully, maybe something run in your dataset

Thanks for your kind help, I don't know why it can run suddenly in the evening, maybe it has some mistake in my dataset or colab(because i train it on colab 0.0)