Skip to content

Latest commit

 

History

History

semi_det

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 
 
 
 
 

简体中文 | English

Semi-Supervised Detection (Semi DET) 半监督检测

内容

简介

半监督目标检测(Semi DET)是同时使用有标注数据和无标注数据进行训练的目标检测,既可以极大地节省标注成本,也可以充分利用无标注数据进一步提高检测精度。PaddleDetection团队提供了DenseTeacherARSL等最前沿的半监督检测算法,用户可以下载使用。

模型库

纯监督数据模型的训练和模型库,请参照Baseline

模型 监督数据比例 Sup Baseline Sup Epochs (Iters) Sup mAPval
0.5:0.95
Semi mAPval
0.5:0.95
Semi Epochs (Iters) 模型下载 配置文件
DenseTeacher-FCOS 5% sup_config 24 (8712) 21.3 30.6 240 (87120) download config
DenseTeacher-FCOS 10% sup_config 24 (17424) 26.3 35.1 240 (174240) download config
DenseTeacher-FCOS(LSJ) 10% sup_config 24 (17424) 26.3 37.1(LSJ) 240 (174240) download config
DenseTeacher-FCOS 100%(full) sup_config 24 (175896) 42.6 44.2 24 (175896) download config
模型 监督数据比例 Sup Baseline Sup Epochs (Iters) Sup mAPval
0.5:0.95
Semi mAPval
0.5:0.95
Semi Epochs (Iters) 模型下载 配置文件
DenseTeacher-PPYOLOE+_s 5% sup_config 80 (14480) 32.8 34.0 200 (36200) download config
DenseTeacher-PPYOLOE+_s 10% sup_config 80 (14480) 35.3 37.5 200 (36200) download config
DenseTeacher-PPYOLOE+_l 5% sup_config 80 (14480) 42.9 45.4 200 (36200) download config
DenseTeacher-PPYOLOE+_l 10% sup_config 80 (14480) 45.7 47.4 200 (36200) download config
模型 COCO监督数据比例 Semi mAPval
0.5:0.95
Semi Epochs (Iters) 模型下载 配置文件
ARSL-FCOS 1% 22.8 240 (87120) download config
ARSL-FCOS 5% 33.1 240 (174240) download config
ARSL-FCOS 10% 36.9 240 (174240) download config
ARSL-FCOS 10% 38.5(LSJ) 240 (174240) download config
ARSL-FCOS full(100%) 45.1 240 (174240) download config

半监督数据集准备

半监督目标检测同时需要有标注数据和无标注数据,且无标注数据量一般远多于有标注数据量。 对于COCO数据集一般有两种常规设置:

(1)抽取部分比例的原始训练集train2017作为标注数据和无标注数据;

train2017中按固定百分比(1%、2%、5%、10%等)抽取,由于抽取方法会对半监督训练的结果影响较大,所以采用五折交叉验证来评估。运行数据集划分制作的脚本如下:

python tools/gen_semi_coco.py

会按照 1%、2%、5%、10% 的监督数据比例来划分train2017全集,为了交叉验证每一种划分会随机重复5次,生成的半监督标注文件如下:

  • 标注数据集标注:instances_train2017.{fold}@{percent}.json
  • 无标注数据集标注:instances_train2017.{fold}@{percent}-unlabeled.json 其中,fold 表示交叉验证,percent 表示有标注数据的百分比。

注意如果根据txt_file生成,需要下载COCO_supervision.txt:

wget https://bj.bcebos.com/v1/paddledet/data/coco/COCO_supervision.txt

(2)使用全量原始训练集train2017作为有标注数据 和 全量原始无标签图片集unlabeled2017作为无标注数据;

下载链接

PaddleDetection团队提供了COCO数据集全部的标注文件,请下载并解压存放至对应目录:

# 下载COCO全量数据集图片和标注
# 包括 train2017, val2017, annotations
wget https://bj.bcebos.com/v1/paddledet/data/coco.tar

# 下载PaddleDetection团队整理的COCO部分比例数据的标注文件
wget https://bj.bcebos.com/v1/paddledet/data/coco/semi_annotations.zip

# unlabeled2017是可选,如果不需要训‘full’则无需下载
# 下载COCO全量 unlabeled 无标注数据集
wget https://bj.bcebos.com/v1/paddledet/data/coco/unlabeled2017.zip
wget https://bj.bcebos.com/v1/paddledet/data/coco/image_info_unlabeled2017.zip
# 下载转换完的 unlabeled2017 无标注json文件
wget https://bj.bcebos.com/v1/paddledet/data/coco/instances_unlabeled2017.zip

如果需要用到COCO全量unlabeled无标注数据集,需要将原版的image_info_unlabeled2017.json进行格式转换,运行以下代码:

COCO unlabeled 标注转换代码:
import json
anns_train = json.load(open('annotations/instances_train2017.json', 'r'))
anns_unlabeled = json.load(open('annotations/image_info_unlabeled2017.json', 'r'))
unlabeled_json = {
  'images': anns_unlabeled['images'],
  'annotations': [],
  'categories': anns_train['categories'],
}
path = 'annotations/instances_unlabeled2017.json'
with open(path, 'w') as f:
  json.dump(unlabeled_json, f)
解压后的数据集目录如下:
PaddleDetection
├── dataset
│   ├── coco
│   │   ├── annotations
│   │   │   ├── instances_train2017.json
│   │   │   ├── instances_unlabeled2017.json
│   │   │   ├── instances_val2017.json
│   │   ├── semi_annotations
│   │   │   ├── [email protected]
│   │   │   ├── [email protected]
│   │   │   ├── [email protected]
│   │   │   ├── [email protected]
│   │   │   ├── [email protected]
│   │   │   ├── [email protected]
│   │   │   ├── [email protected]
│   │   │   ├── [email protected]
│   │   ├── train2017
│   │   ├── unlabeled2017
│   │   ├── val2017

半监督检测配置

配置半监督检测,需要基于选用的基础检测器的配置文件,如:

_BASE_: [
  '../../fcos/fcos_r50_fpn_iou_multiscale_2x_coco.yml',
  '../_base_/coco_detection_percent_10.yml',
]
log_iter: 50
snapshot_epoch: 5
epochs: &epochs 240
weights: output/denseteacher_fcos_r50_fpn_coco_semi010/model_final

并依次做出如下几点改动:

训练集配置

首先可以直接引用已经配置好的半监督训练集,如:

_BASE_: [
  '../_base_/coco_detection_percent_10.yml',
]

具体来看,构建半监督数据集,需要同时配置监督数据集TrainDataset和无监督数据集UnsupTrainDataset的路径,注意必须选用SemiCOCODataSet类而不是COCODataSet,如以下所示:

COCO-train2017部分比例数据集

# partial labeled COCO, use `SemiCOCODataSet` rather than `COCODataSet`
TrainDataset:
  !SemiCOCODataSet
    image_dir: train2017
    anno_path: semi_annotations/instances_train2017.1@10.json
    dataset_dir: dataset/coco
    data_fields: ['image', 'gt_bbox', 'gt_class', 'is_crowd']

# partial unlabeled COCO, use `SemiCOCODataSet` rather than `COCODataSet`
UnsupTrainDataset:
  !SemiCOCODataSet
    image_dir: train2017
    anno_path: semi_annotations/instances_train2017.1@10-unlabeled.json
    dataset_dir: dataset/coco
    data_fields: ['image']
    supervised: False

或者 COCO-train2017 full全量数据集

# full labeled COCO, use `SemiCOCODataSet` rather than `COCODataSet`
TrainDataset:
  !SemiCOCODataSet
    image_dir: train2017
    anno_path: annotations/instances_train2017.json
    dataset_dir: dataset/coco
    data_fields: ['image', 'gt_bbox', 'gt_class', 'is_crowd']

# full unlabeled COCO, use `SemiCOCODataSet` rather than `COCODataSet`
UnsupTrainDataset:
  !SemiCOCODataSet
    image_dir: unlabeled2017
    anno_path: annotations/instances_unlabeled2017.json
    dataset_dir: dataset/coco
    data_fields: ['image']
    supervised: False

验证集EvalDataset和测试集TestDataset的配置不需要更改,且还是采用COCODataSet类。

预训练配置

### pretrain and warmup config, choose one and comment another
pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/ResNet50_cos_pretrained.pdparams
semi_start_iters: 5000
ema_start_iters: 3000
use_warmup: &use_warmup True

注意:

  • Dense Teacher原文使用R50-va-caffe预训练,PaddleDetection中默认使用R50-vb预训练,如果使用R50-vd结合SSLD的预训练模型,可进一步显著提升检测精度,同时backbone部分配置也需要做出相应更改,如:
 pretrain_weights:  https://paddledet.bj.bcebos.com/models/pretrained/ResNet50_vd_ssld_v2_pretrained.pdparams
 ResNet:
   depth: 50
   variant: d
   norm_type: bn
   freeze_at: 0
   return_idx: [1, 2, 3]
   num_stages: 4
   lr_mult_list: [0.05, 0.05, 0.1, 0.15]

全局配置

需要在配置文件中添加如下全局配置,并且注意 DenseTeacher 模型需要使用use_simple_ema: True而不是use_ema: True

### global config
use_simple_ema: True
ema_decay: 0.9996
ssod_method: DenseTeacher
DenseTeacher:
  train_cfg:
    sup_weight: 1.0
    unsup_weight: 1.0
    loss_weight: {distill_loss_cls: 4.0, distill_loss_box: 1.0, distill_loss_quality: 1.0}
    concat_sup_data: True
    suppress: linear
    ratio: 0.01
    gamma: 2.0
  test_cfg:
    inference_on: teacher

模型配置

如果没有特殊改动,则直接继承自基础检测器里的模型配置。 以 DenseTeacher 为例,选择 fcos_r50_fpn_iou_multiscale_2x_coco.yml 作为基础检测器进行半监督训练,teacher网络的结构和student网络的结构均为基础检测器的结构,且结构相同

_BASE_: [
  '../../fcos/fcos_r50_fpn_iou_multiscale_2x_coco.yml',
]

数据增强配置

构建半监督训练集的Reader,需要在原先TrainReader的基础上,新增加weak_aug,strong_aug,sup_batch_transformsunsup_batch_transforms,并且需要注意:

  • 如果有NormalizeImage,需要单独从sample_transforms中抽出来放在weak_augstrong_aug中;
  • sample_transforms公用的基础数据增强
  • 完整的弱数据增强为sample_transforms + weak_aug,完整的强数据增强为sample_transforms + strong_aug

如以下所示:

原纯监督模型的TrainReader

TrainReader:
  sample_transforms:
    - Decode: {}
    - RandomResize: {target_size: [[640, 1333], [672, 1333], [704, 1333], [736, 1333], [768, 1333], [800, 1333]], keep_ratio: True, interp: 1}
    - RandomFlip: {}
    - NormalizeImage: {mean: [0.485, 0.456, 0.406], std: [0.229, 0.224, 0.225], is_scale: True}
  batch_transforms:
    - Permute: {}
    - PadBatch: {pad_to_stride: 32}
    - Gt2FCOSTarget:
        object_sizes_boundary: [64, 128, 256, 512]
        center_sampling_radius: 1.5
        downsample_ratios: [8, 16, 32, 64, 128]
        norm_reg_targets: True
  batch_size: 2
  shuffle: True
  drop_last: True

更改后的半监督TrainReader:

### reader config
SemiTrainReader:
  sample_transforms:
    - Decode: {}
    - RandomResize: {target_size: [[640, 1333], [672, 1333], [704, 1333], [736, 1333], [768, 1333], [800, 1333]], keep_ratio: True, interp: 1}
    - RandomFlip: {}
  weak_aug:
    - NormalizeImage: {mean: [0.485, 0.456, 0.406], std: [0.229, 0.224, 0.225], is_scale: true}
  strong_aug:
    - StrongAugImage: {transforms: [
        RandomColorJitter: {prob: 0.8, brightness: 0.4, contrast: 0.4, saturation: 0.4, hue: 0.1},
        RandomErasingCrop: {},
        RandomGaussianBlur: {prob: 0.5, sigma: [0.1, 2.0]},
        RandomGrayscale: {prob: 0.2},
      ]}
    - NormalizeImage: {mean: [0.485, 0.456, 0.406], std: [0.229, 0.224, 0.225], is_scale: true}
  sup_batch_transforms:
    - Permute: {}
    - PadBatch: {pad_to_stride: 32}
    - Gt2FCOSTarget:
        object_sizes_boundary: [64, 128, 256, 512]
        center_sampling_radius: 1.5
        downsample_ratios: [8, 16, 32, 64, 128]
        norm_reg_targets: True
  unsup_batch_transforms:
    - Permute: {}
    - PadBatch: {pad_to_stride: 32}
  sup_batch_size: 2
  unsup_batch_size: 2
  shuffle: True
  drop_last: True

其他配置

训练epoch数需要和全量数据训练时换算总iter数保持一致,如全量训练24 epoch(换算约为180k个iter),则10%监督数据的半监督训练,总epoch数需要为240 epoch左右(换算约为180k个iter)。示例如下:

### other config
epoch: 240
LearningRate:
  base_lr: 0.01
  schedulers:
  - !PiecewiseDecay
    gamma: 0.1
    milestones: 240
    use_warmup: True
  - !LinearWarmup
    start_factor: 0.001
    steps: 1000

OptimizerBuilder:
  optimizer:
    momentum: 0.9
    type: Momentum
  regularizer:
    factor: 0.0001
    type: L2
  clip_grad_by_value: 1.0

使用说明

仅训练时必须使用半监督检测的配置文件去训练,评估、预测、部署也可以按基础检测器的配置文件去执行。

训练

# 单卡训练 (不推荐,需按线性比例相应地调整学习率)
CUDA_VISIBLE_DEVICES=0 python tools/train.py -c configs/semi_det/denseteacher/denseteacher_fcos_r50_fpn_coco_semi010.yml --eval

# 多卡训练
python -m paddle.distributed.launch --log_dir=denseteacher_fcos_semi010/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/semi_det/denseteacher/denseteacher_fcos_r50_fpn_coco_semi010.yml --eval

评估

CUDA_VISIBLE_DEVICES=0 python tools/eval.py -c configs/semi_det/denseteacher/denseteacher_fcos_r50_fpn_coco_semi010.yml -o weights=output/denseteacher_fcos_r50_fpn_coco_semi010/model_final.pdparams

预测

CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/semi_det/denseteacher/denseteacher_fcos_r50_fpn_coco_semi010.yml -o weights=output/denseteacher_fcos_r50_fpn_coco_semi010/model_final.pdparams --infer_img=demo/000000014439.jpg

部署

部署可以使用半监督检测配置文件,也可以使用基础检测器的配置文件去部署和使用。

# 导出模型
CUDA_VISIBLE_DEVICES=0 python tools/export_model.py -c configs/semi_det/denseteacher/denseteacher_fcos_r50_fpn_coco_semi010.yml -o weights=https://paddledet.bj.bcebos.com/models/denseteacher_fcos_r50_fpn_coco_semi010.pdparams

# 导出权重预测
CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/denseteacher_fcos_r50_fpn_coco_semi010 --image_file=demo/000000014439_640x640.jpg --device=GPU

# 部署测速
CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/denseteacher_fcos_r50_fpn_coco_semi010 --image_file=demo/000000014439_640x640.jpg --device=GPU --run_benchmark=True # --run_mode=trt_fp16

# 导出ONNX
paddle2onnx --model_dir output_inference/denseteacher_fcos_r50_fpn_coco_semi010/ --model_filename model.pdmodel --params_filename model.pdiparams --opset_version 12 --save_file denseteacher_fcos_r50_fpn_coco_semi010.onnx

引用

 @article{denseteacher2022,
  title={Dense Teacher: Dense Pseudo-Labels for Semi-supervised Object Detection},
  author={Hongyu Zhou, Zheng Ge, Songtao Liu, Weixin Mao, Zeming Li, Haiyan Yu, Jian Sun},
  journal={arXiv preprint arXiv:2207.02541},
  year={2022}
}