nanodet训练voc数据集

首先是修改配置文件，config/nanodet_custom_xml_dataset.yml

主要是修改以下几项
 

class_names: &class_names ['aeroplane','bicycle','bird','boat','bottle','bus','car','cat','chair','cow','diningtable',
'dog','horse','motorbike','person','pottedplant','sheep','sofa','train','tvmonitor']

num_classes: 20

img_path: /media/sf_D_DRIVE/tmp/data/VOCtrainval_11-May-2012/VOCdevkit/VOC2012/JPEGImages
ann_path: /media/sf_D_DRIVE/tmp/data/VOCtrainval_11-May-2012/VOCdevkit/VOC2012/Annotations

参考全部文件

#Config File example
save_dir: workspace/nanodet_m
model:
  weight_averager:
    name: ExpMovingAverager
    decay: 0.9998
  arch:
    name: NanoDetPlus
    detach_epoch: 10
    backbone:
      name: ShuffleNetV2
      model_size: 1.0x
      out_stages: [2,3,4]
      activation: LeakyReLU
    fpn:
      name: GhostPAN
      in_channels: [116, 232, 464]
      out_channels: 96
      kernel_size: 5
      num_extra_level: 1
      use_depthwise: True
      activation: LeakyReLU
    head:
      name: NanoDetPlusHead
      num_classes: 20
      input_channel: 96
      feat_channels: 96
      stacked_convs: 2
      kernel_size: 5
      strides: [8, 16, 32, 64]
      activation: LeakyReLU
      reg_max: 7
      norm_cfg:
        type: BN
      loss:
        loss_qfl:
          name: QualityFocalLoss
          use_sigmoid: True
          beta: 2.0
          loss_weight: 1.0
        loss_dfl:
          name: DistributionFocalLoss
          loss_weight: 0.25
        loss_bbox:
          name: GIoULoss
          loss_weight: 2.0
    # Auxiliary head, only use in training time.
    aux_head:
      name: SimpleConvHead
      num_classes: 20
      input_channel: 192
      feat_channels: 192
      stacked_convs: 4
      strides: [8, 16, 32, 64]
      activation: LeakyReLU
      reg_max: 7

class_names: &class_names ['aeroplane','bicycle','bird','boat','bottle','bus','car','cat','chair','cow','diningtable',
'dog','horse','motorbike','person','pottedplant','sheep','sofa','train','tvmonitor']
data:
  train:
    name: XMLDataset
    class_names: *class_names
    img_path: /media/sf_D_DRIVE/tmp/data/VOCtrainval_11-May-2012/VOCdevkit/VOC2012/JPEGImages 
    ann_path: /media/sf_D_DRIVE/tmp/data/VOCtrainval_11-May-2012/VOCdevkit/VOC2012/Annotations
    input_size: [320,320] #[w,h]
    keep_ratio: True
    pipeline:
      perspective: 0.0
      scale: [0.6, 1.4]
      stretch: [[1, 1], [1, 1]]
      rotation: 0
      shear: 0
      translate: 0.2
      flip: 0.5
      brightness: 0.2
      contrast: [0.8, 1.2]
      saturation: [0.8, 1.2]
      normalize: [[103.53, 116.28, 123.675], [57.375, 57.12, 58.395]]
  val:
    name: XMLDataset
    class_names: *class_names
    img_path: /media/sf_D_DRIVE/tmp/data/VOCtrainval_11-May-2012/VOCdevkit/VOC2012/JPEGImages #Please fill in val image path
    ann_path: /media/sf_D_DRIVE/tmp/data/VOCtrainval_11-May-2012/VOCdevkit/VOC2012/Annotations #Please fill in val xml path
    input_size: [320,320] #[w,h]
    keep_ratio: True
    pipeline:
      normalize: [[103.53, 116.28, 123.675], [57.375, 57.12, 58.395]]
device:
  gpu_ids: [0] # Set like [0, 1, 2, 3] if you have multi-GPUs
  workers_per_gpu: 8
  batchsize_per_gpu: 96
schedule:
#  resume:
#  load_model: YOUR_MODEL_PATH
  optimizer:
    name: AdamW
    lr: 0.001
    weight_decay: 0.05
  warmup:
    name: linear
    steps: 500
    ratio: 0.0001
  total_epochs: 300
  lr_schedule:
    name: CosineAnnealingLR
    T_max: 300
    eta_min: 0.00005
  val_intervals: 10
grad_clip: 35
evaluator:
  name: CocoDetectionEvaluator
  save_key: mAP

log:
  interval: 10

开始训练

export PYTHONPATH=.
python tools/train.py config/nanodet_custom_xml_dataset.yml

voc的数据集的标签文件，Annotations/2011_003353.xml的ymin是一个float类型，运行时代码会报错，

ValueError: invalid literal for int() with base 10:

我就没改代码，直接把45.7改为了45

同样的还有2011_006777.xml

当然上面的配置是比较暴力,把所有数据都作为训练数据和验证数据,这肯定是不合理的

那我们可以把一部分数据移动到另一个文件夹即可,如

mv *9.xml ../test/Annotations

然后把验证数据的 ann_path 设定为新建的文件夹即可