标注的数据集在上一篇文章介绍过了,先使用labelme软件标注好,然后使用代码,生成一个汇总的json文件
http://t.csdnimg.cn/oKqDD
(mmdet300rc5_torch171) PS G:\bsh\mmdet300rc5_torch171> python
Python 3.8.18 (default, Sep 11 2023, 13:39:12) [MSC v.1916 64 bit (AMD64)] :: Anaconda, Inc. on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> import torchvision
>>> torch.__version__
'1.7.1'
>>> torchvision.__version__
'0.8.2'
>>> torch.version.cuda
'10.1'
>>> torch.cuda.get_device_name(0)
'NVIDIA GeForce GTX 1060'
>>> quit()
(mmdet300rc5_torch171) PS G:\bsh\mmdet300rc5_torch171> mim list
Package Version Source
--------- --------- ----------------------------------------------------------------------------
mmcv 2.0.0rc4 https://github.com/open-mmlab/mmcv
mmdet 3.2.0 f:\app\miniconda\envs\mmdet_torch171\lib\site-packages\mmdet-3.2.0-py3.8.egg
mmengine 0.9.0 https://github.com/open-mmlab/mmengine
G:\bsh\mmdet300rc5_torch171\data\flame_coco\annotations\instances_train.json
G:\bsh\mmdet300rc5_torch171\data\flame_coco\annotations\instances_val.json
G:\bsh\mmdet300rc5_torch171\data\flame_coco\images\train
G:\bsh\mmdet300rc5_torch171\data\flame_coco\images\val
num_classes=2、
batch_size=12、
num_workers=1(这个必须改为1,windows不支持多线程训练)、
max_epochs=30、
base_lr=0.004、
interval=10、
pin_memory=False,(这个必须改为False,windows不支持多线程训练)
@DATASETS.register_module()
class CocoDataset(BaseDetDataset):
"""Dataset for COCO."""
METAINFO = {
'classes':
('ok2', 'ok10'),
# palette is a list of color tuples, which is used for visualization.
'palette':
[(220, 20, 60), (119, 11, 32), (0, 0, 142), (0, 0, 230), (106, 0, 228),
(0, 60, 100), (0, .......................
def coco_classes():
return [
'ok2', 'ok10'
]
# dataset settings
dataset_type = 'CocoDataset'
data_root = 'data/flame_coco/'
# file_client_args = dict(
# backend='petrel',
# path_mapping=dict({
# './data/': 's3://openmmlab/datasets/detection/',
# 'data/': 's3://openmmlab/datasets/detection/'
# }))
file_client_args = dict(backend='disk')
train_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadAnnotations', with_bbox=True),
dict(type='Resize', scale=(1333, 800), keep_ratio=True),
dict(type='RandomFlip', prob=0.5),
dict(type='PackDetInputs')
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='Resize', scale=(1333, 800), keep_ratio=True),
# If you don't have a gt annotation, delete the pipeline
dict(type='LoadAnnotations', with_bbox=True),
dict(
type='PackDetInputs',
meta_keys=('img_id', 'img_path', 'ori_shape', 'img_shape',
'scale_factor'))
]
train_dataloader = dict(
batch_size=2,
num_workers=2,
persistent_workers=True,
sampler=dict(type='DefaultSampler', shuffle=True),
batch_sampler=dict(type='AspectRatioBatchSampler'),
dataset=dict(
type=dataset_type,
data_root=data_root,
ann_file='annotations/instances_train.json',
data_prefix=dict(img='images/train/'),
filter_cfg=dict(filter_empty_gt=True, min_size=32),
pipeline=train_pipeline))
val_dataloader = dict(
batch_size=1,
num_workers=2,
persistent_workers=True,
drop_last=False,
sampler=dict(type='DefaultSampler', shuffle=False),
dataset=dict(
type=dataset_type,
data_root=data_root,
ann_file='annotations/instances_val.json',
data_prefix=dict(img='images/val/'),
test_mode=True,
pipeline=test_pipeline))
test_dataloader = val_dataloader
val_evaluator = dict(
type='CocoMetric',
ann_file=data_root + 'annotations/instances_val.json',
metric='bbox',
format_only=False)
test_evaluator = val_evaluator
python tools/train.py configs\rtmdet\rtmdet_s_8xb32-300e_coco.py
python demo/image_demo.py demo/00000.bmp configs/rtmdet/rtmdet_s_8xb32-300e_coco.py --weights work_dirs/rtmdet_s_8xb32-300e_coco/epoch_300.pth --show