Use MMDetection to train your own data set (Colab version)

user's guidance

Mount Google Cloud Disk

from google.colab import drive
drive.mount('/content/drive')

Confirm connection to GPU

import tensorflow as tf

device_name = tf.test.gpu_device_name()
if device_name != '/device:GPU:0':
  raise SystemError('没有发现GPU device')
print('Found GPU at: {}'.format(device_name))
# Found GPU at: /device:GPU:0

Check the graphics card

!/opt/bin/nvidia-smi

Install OpenMMLab dependencies

!pip install openmim
!mim install mmdet

Check if the dependencies are installed

from mmcv.runner import checkpoint  
# 测试语句
from mmdet.apis import inference_detector, init_detector, show_result_pyplot
print("载入成功！")

switch working directory

import os
os.chdir("/content/drive/MyDrive/mmdetection")
os.getcwd()

modify the number of categories

修改 mmdet/core/evaluation/class_names.py，return自己的类别
修改 mmdet/datasets/coco.py，将 CLASSES = () 修改成自己的类别。

recompile

!python setup.py install

My yolact_r50_1x8_coco.pyconfiguration file is as follows

_base_ = '../_base_/default_runtime.py'

# model settings
img_size = 550
# num_classes = 11
checkpoint_config = dict(   # Checkpoint hook 的配置文件。执行时请参考 https://github.com/open-mmlab/mmcv/blob/master/mmcv/runner/hooks/checkpoint.py。
    interval=5)  # 保存的间隔是 5。
model = dict(
    type='YOLACT',
    backbone=dict(
        type='ResNet',
        depth=50,
        num_stages=4,
        out_indices=(0, 1, 2, 3),
        frozen_stages=-1,  # do not freeze stem
        norm_cfg=dict(type='BN', requires_grad=True),
        norm_eval=False,  # update the statistics of bn
        zero_init_residual=False,
        style='pytorch',
        init_cfg=dict(type='Pretrained', checkpoint='torchvision://resnet50')),
    neck=dict(
        type='FPN',
        in_channels=[256, 512, 1024, 2048],
        out_channels=256,
        start_level=1,
        add_extra_convs='on_input',
        num_outs=5,
        upsample_cfg=dict(mode='bilinear')),
    bbox_head=dict(
        type='YOLACTHead',
        num_classes=11,
        in_channels=256,
        feat_channels=256,
        anchor_generator=dict(
            type='AnchorGenerator',
            octave_base_scale=3,
            scales_per_octave=1,
            base_sizes=[8, 16, 32, 64, 128],
            ratios=[0.5, 1.0, 2.0],
            strides=[550.0 / x for x in [69, 35, 18, 9, 5]],
            centers=[(550 * 0.5 / x, 550 * 0.5 / x)
                     for x in [69, 35, 18, 9, 5]]),
        bbox_coder=dict(
            type='DeltaXYWHBBoxCoder',
            target_means=[.0, .0, .0, .0],
            target_stds=[0.1, 0.1, 0.2, 0.2]),
        loss_cls=dict(
            type='CrossEntropyLoss',
            use_sigmoid=False,
            reduction='none',
            loss_weight=1.0),
        loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.5),
        num_head_convs=1,
        num_protos=32,
        use_ohem=True),
    mask_head=dict(
        type='YOLACTProtonet',
        in_channels=256,
        num_protos=32,
        num_classes=11,
        max_masks_to_train=100,
        loss_mask_weight=6.125),
    segm_head=dict(
        type='YOLACTSegmHead',
        num_classes=11,
        in_channels=256,
        loss_segm=dict(
            type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0)),
    # training and testing settings
    train_cfg=dict(
        assigner=dict(
            type='MaxIoUAssigner',
            pos_iou_thr=0.5,
            neg_iou_thr=0.4,
            min_pos_iou=0.,
            ignore_iof_thr=-1,
            gt_max_assign_all=False),
        # smoothl1_beta=1.,
        allowed_border=-1,
        pos_weight=-1,
        neg_pos_ratio=3,
        debug=False),
    test_cfg=dict(
        nms_pre=1000,
        min_bbox_size=0,
        score_thr=0.05,
        iou_thr=0.5,
        top_k=200,
        max_per_img=100))
# dataset settings
dataset_type = 'CocoDataset'
# declare the classes name
classes = ('Squamous','WBC','Urothelial','UKA','CaOX','RBC','Hyaline','Granular','UA','YEAST','Renal')
data_root = 'data/coco/'
img_norm_cfg = dict(
    mean=[123.68, 116.78, 103.94], std=[58.40, 57.12, 57.38], to_rgb=True)
train_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(type='LoadAnnotations', with_bbox=True, with_mask=True),
    dict(type='FilterAnnotations', min_gt_bbox_wh=(4.0, 4.0)),
    dict(
        type='Expand',
        mean=img_norm_cfg['mean'],
        to_rgb=img_norm_cfg['to_rgb'],
        ratio_range=(1, 4)),
    dict(
        type='MinIoURandomCrop',
        min_ious=(0.1, 0.3, 0.5, 0.7, 0.9),
        min_crop_size=0.3),
    dict(type='Resize', img_scale=(img_size, img_size), keep_ratio=False),
    dict(type='RandomFlip', flip_ratio=0.5),
    dict(
        type='PhotoMetricDistortion',
        brightness_delta=32,
        contrast_range=(0.5, 1.5),
        saturation_range=(0.5, 1.5),
        hue_delta=18),
    dict(type='Normalize', **img_norm_cfg),
    dict(type='DefaultFormatBundle'),
    dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels', 'gt_masks']),
]
test_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(
        type='MultiScaleFlipAug',
        img_scale=(img_size, img_size),
        flip=False,
        transforms=[
            dict(type='Resize', keep_ratio=False),
            dict(type='Normalize', **img_norm_cfg),
            dict(type='ImageToTensor', keys=['img']),
            dict(type='Collect', keys=['img']),
        ])
]
data = dict(
    samples_per_gpu=8,
    workers_per_gpu=4,
    train=dict(
        classes = classes,
        type=dataset_type,
        ann_file=data_root + 'annotations/train325.json',
        img_prefix=data_root + 'train_img/',
        pipeline=train_pipeline),
    val=dict(
        classes = ('Squamous','WBC','Urothelial','UKA','CaOX','RBC','Hyaline','Granular','UA','YEAST','Renal'),
        type=dataset_type,
        ann_file=data_root + 'annotations/test134.json',
        img_prefix=data_root + 'test_img/',
        pipeline=test_pipeline),
    test=dict(
        classes = ('Squamous','WBC','Urothelial','UKA','CaOX','RBC','Hyaline','Granular','UA','YEAST','Renal'),
        type=dataset_type,
        ann_file=data_root + 'annotations/test134.json',
        img_prefix=data_root + 'test_img/',
        pipeline=test_pipeline))
# optimizer
optimizer = dict(type='SGD', lr=1e-3, momentum=0.9, weight_decay=5e-4)
optimizer_config = dict()
# learning policy
lr_config = dict(
    policy='step',
    warmup='linear',
    warmup_iters=500,
    warmup_ratio=0.1,
    step=[20, 42, 49, 52])
runner = dict(type='EpochBasedRunner', max_epochs=20)
cudnn_benchmark = True
evaluation = dict(metric=['bbox', 'segm'])

Run python file training

!python  tools/train.py  --auto-resume pig_work_dir/yolact/yolact_r50_1x8_coco.py

test statement

from mmcv.runner import checkpoint  
# 测试语句
from mmdet.apis import inference_detector, init_detector, show_result_pyplot

# print("载入成功！")
# 模型配置文件
config = '/content/drive/MyDrive/mmdetection/work_dirs/yolact_r50_1x8_coco/yolact_r50_1x8_coco.py'
# 模型文件
checkpoint = '/content/drive/MyDrive/mmdetection/work_dirs/yolact_r50_1x8_coco/latest.pth'
# 初始化检测器
model = init_detector(config, checkpoint, device='cuda:0')

# 使用检测器去预测
img = '/content/drive/MyDrive/mmdetection/data/coco/test_img/bd000005.jpg'
result = inference_detector(model, img)

# 查看结果
show_result_pyplot(model, img, result, score_thr=0.3)

Prevent colab from dropping (put this code in the console)

function ClickConnect(){
    
    
  console.log("Working"); 
  document
    .querySelector("#top-toolbar > colab-connect-button")
    .shadowRoot
    .querySelector("#connect")
    .click()
}
 
var id=setInterval(ClickConnect,5*60000)   //5分钟点一次，改变频率把5换成其他数即可，单位分钟
 
 
//要提前停止，请输入运行以下代码：    clearInterval(id)

report error

error 1

AssertionError: The num_classes (3) in Shared2FCBBoxHead of MMDataParallel does not matches the length of CLASSES 80) in CocoDataset

Solution

Add category information to config file

data = dict(
train=dict(
classes=('Squamous','WBC','Urothelial','UKA','CaOX','RBC','Hyaline','Granular','UA','YEAST' ,'Renal'), # your own category type=dataset_type,

​ …

))

details as follows

data = dict(
    samples_per_gpu=8,
    workers_per_gpu=4,
    train=dict(
        classes = ('Squamous','WBC','Urothelial','UKA','CaOX','RBC','Hyaline','Granular','UA','YEAST','Renal'),
        type=dataset_type,
        ann_file=data_root + 'annotations/train325.json',
        img_prefix=data_root + 'train_img/',
        pipeline=train_pipeline),
    val=dict(
        classes = ('Squamous','WBC','Urothelial','UKA','CaOX','RBC','Hyaline','Granular','UA','YEAST','Renal'),
        type=dataset_type,
        ann_file=data_root + 'annotations/test134.json',
        img_prefix=data_root + 'test_img/',
        pipeline=test_pipeline),
    test=dict(
        classes = ('Squamous','WBC','Urothelial','UKA','CaOX','RBC','Hyaline','Granular','UA','YEAST','Renal'),
        type=dataset_type,
        ann_file=data_root + 'annotations/test134.json',
        img_prefix=data_root + 'test_img/',
        pipeline=test_pipeline))

Finally, thank you for your study~