[1804.06208] Simple Baselines for Human Pose Estimation and Tracking (arxiv.org)
[1603.06937] Stacked Hourglass Networks for Human Pose Estimation (arxiv.org)
-
安装python>=3.6; pytorch >= v1.0.0
-
安装依赖:
pip install -r requirements.txt
-
安装 COCOAPI:
# COCOAPI=/path/to/clone/cocoapi git clone https://github.com/cocodataset/cocoapi.git $COCOAPI cd $COCOAPI/PythonAPI # Install into global site-packages make install # Alternatively, if you do not have permissions or prefer # not to install the COCO API into global site-packages python3 setup.py install --user
Note that instructions like # COCOAPI=/path/to/install/cocoapi indicate that you should pick a path where you'd like to have the software cloned and then set an environment variable (COCOAPI in this case) accordingly.
-
下载Pytorch resnet预训练模型.
${POSE_ROOT} `-- models `-- pytorch |-- imagenet | |-- resnet50-19c8e357.pth | |-- resnet101-5d3b4d8f.pth | `-- resnet152-b121ed2d.pth `-- trained
-
Init output(training model output directory) and log(tensorboard log directory) directory:
mkdir output mkdir log
Your directory tree should look like this:
${POSE_ROOT} ├── data │ └── fish │ ├── annotations │ │ └── fish_7.11_coco.json │ └── images │ └── fish22.6.17 ├── experiments ├── lib ├── log ├── models ├── output ├── pose_estimation ├── README.md └── requirements.txt
${POSE_ROOT}
└── data
└── fish
├── annotations
│ └── fish_7.11_coco.json
└── images
└── fish22.6.17experiments/coco/
DATASET:
DATASET: 'coco'
ROOT: 'data/fish/' #数据集路径
FOLDER: 'fish22.6.17' #图片文件夹
TEST_SET: 'fish_blank_2400' #测试label文件名（coco格式）
TRAIN_SET: 'fish_7.11_coco' #训练label文件名（coco格式）MODEL:
NAME: 'pose_resnet' #可选['pose_renet', 'hourglass']
NUM_JOINTS: 9 #点的数量
N_STACK: 8 #Hourglass堆叠数量TEST:
MODEL_FILE: 'models/pytorch/trained/final_state.pth.tar' #保存的权重文件路径其余的训练参数均可自行调节
将自己的标签转化成coco格式，预测使用的json可以没有关键点坐标数据，只需要有bbox(检测框)
{
"images": [
{
"file_name": "IMG_20220318_093630.jpg",
"id": 1, #图片id，每张图片需要不同
"height": 5616, #图片大小
"width": 2592
},
·
·
·
],
"annotations": [
{
"iscrowd": 0,
"bbox": [
1053.99, #检测框左上坐标x
1056.258, #坐标y
1109.7430000000002, #宽width
4851.512000000001 #高height
],
"image_id": 1, #图片id
"category_id": 1,
"id": 1,
"area": 5383931.481416002, #宽×高
"keypoints": [ #关键点坐标，若点在图片外第三维为0，
1525.89, 若为预测标签，第三维全为2，点坐标为0
1173.62,
2,
1662.9,
2066.9,
2,
1453.1,
2058.0,
2,
1967.03,
·
·
·
],
"num_keypoints": 9 #点的数量
},
·
·
·
·
],
"categories": [
{
"supercategory": "person",
"id": 1,
"name": "person",
"keypoints": [ #修改点名称，其余直接复制
"Eh",
"Et"
],
"skeleton": [
[
16,
14
],
[
14,
12
],
[
17,
15
],
[
15,
13
],
[
12,
13
],
[
6,
12
],
[
7,
13
],
[
6,
7
],
[
6,
8
],
[
7,
9
],
[
8,
10
],
[
9,
11
],
[
2,
3
],
[
1,
2
],
[
1,
3
],
[
2,
4
],
[
3,
5
],
[
4,
6
],
[
5,
7
]
]
}
]
}SimpleBaseline
python pose_estimation/train.py \
--cfg experiments/coco/resnet152/384x384_fish.yamlStackedHourglass
python pose_estimation/trainHG.py \
--cfg experiments/coco/hourglass/4stacks_fish.yamlSimpleBaseline
python pose_estimation/predict.py \
--cfg experiments/coco/resnet152/384x384_fish.yamlStackedHourglass
python pose_estimation/predictHG.py \
--cfg experiments/coco/hourglass/4stacks_fish.yaml自行修改yaml路径
训练和预测的结果保存在output文件夹中