Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/deepvac/dbnet
DeepVAC-compliant DB Net implementation.
https://github.com/deepvac/dbnet
Last synced: about 2 months ago
JSON representation
DeepVAC-compliant DB Net implementation.
- Host: GitHub
- URL: https://github.com/deepvac/dbnet
- Owner: DeepVAC
- License: gpl-3.0
- Created: 2021-04-13T02:19:53.000Z (over 3 years ago)
- Default Branch: master
- Last Pushed: 2021-06-30T07:29:31.000Z (over 3 years ago)
- Last Synced: 2023-03-05T16:49:49.364Z (almost 2 years ago)
- Language: Python
- Size: 80.1 KB
- Stars: 10
- Watchers: 2
- Forks: 4
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# DBNet
DeepVAC-compliant DBNet implementation.
# 简介
本项目实现了符合DeepVAC规范的OCR检测模型DBNet
**项目依赖**
- deepvac >= 0.5.6
- pytorch >= 1.8.0
- torchvision >= 0.7.0
- opencv-python
- numpy
- pyclipper
- shapely
- pillow
# 如何运行本项目
## 1. 阅读[DeepVAC规范](https://github.com/DeepVAC/deepvac)
可以粗略阅读，建立起第一印象
## 2. 准备运行环境
可以使用DeepVAC规范指定的[Docker镜像](https://github.com/DeepVAC/deepvac#2-%E7%8E%AF%E5%A2%83%E5%87%86%E5%A4%87)
## 3. 准备数据集
- 获取文本检测数据集
CTW1500格式的数据集，CTW1500下载地址:
[ch4_training_images.zip](https://rrc.cvc.uab.es/downloads/ch4_training_images.zip)
[ch4_training_localization_transcription_gt.zip](https://rrc.cvc.uab.es/downloads/ch4_training_localization_transcription_gt.zip)
[ch4_test_images.zip](https://rrc.cvc.uab.es/downloads/ch4_test_images.zip)
[Challenge4_Test_Task1_GT.zip](https://rrc.cvc.uab.es/downloads/Challenge4_Test_Task1_GT.zip)
- 数据集配置
在config.py文件中作如下配置:
```python
config.sample_path =
config.label_path =
config.sample_path =
config.label_path =
```
## 5. 模型相关配置
- DB backbone配置
```python
# 目前支持resnet18，mv3large
config.arch = "resnet18"
```
## 4. 训练相关配置
- dataloader相关配置
```python
config.is_transform = True # 是否做数据增强
config.img_size = 640 # 训练图片大小(img_size, img_size)
config.datasets.DBTrainDataset = AttrDict()
config.datasets.DBTrainDataset.shrink_ratio = 0.4
config.datasets.DBTrainDataset.thresh_min = 0.3
config.datasets.DBTrainDataset.thresh_max = 0.7
config.core.DBNetTrain.batch_size = 8
config.core.DBNetTrain.num_workers = 4
config.core.DBNetTrain.train_dataset = DBTrainDataset(config, config.sample_path, config.label_path, config.is_transform, config.img_size)
config.core.DBNetTrain.train_loader = torch.utils.data.DataLoader(
dataset = config.core.DBNetTrain.train_dataset,
batch_size = config.core.DBNetTrain.batch_size,
shuffle = True,
num_workers = config.core.DBNetTrain.num_workers,
pin_memory = True,
sampler = None
)
```
## 5. 训练
```
python3 train.py
```
## 6. 测试
- 测试相关配置
```python
config.core.DBNetTest.model_path = # 加载模型路径
# config.core.DBNetTest.jit_model_path = # torchscript model path
config.core.DBNetTest.is_output_polygon = True # 输出是否为多边形模型
config.sample_path = # 测试图片路径
config.core.DBNetTrain.batch_size = 8
config.core.DBNetTrain.num_workers = 4
config.core.DBNetTest.test_dataset = DBTestDataset(config, config.sample_path, long_size = 1280)
config.core.DBNetTest.test_loader = torch.utils.data.DataLoader(
dataset = config.core.DBNetTest.test_dataset,
batch_size = config.core.DBNetTrain.batch_size,
shuffle = False,
num_workers = config.core.DBNetTrain.num_workers,
pin_memory = True
)
```
- 运行测试脚本:
```bash
python3 test.py
```
## 7. 使用torchscript模型
如果训练过程中未开启config.cast.TraceCast.model_dir开关，可以在测试过程中转化torchscript模型
- 转换torchscript模型(.pt)
```python
config.cast.TraceCast.model_dir = "output/script.pt"
```
按照步骤6完成测试，torchscript模型会保存至config.cast.TraceCast.model_dir指定位置
- 加载torchscript模型
```python
config.core.DBNetTest.jit_model_path =
```
然后按照步骤6测试，会读取script_model
## 更多功能
如果要在本项目中开启如下功能:
- 预训练模型加载
- checkpoint加载
- 使用tensorboard
- 启用TorchScript
- 转换ONNX
- 转换NCNN
- 转换CoreML
- 开启量化
- 开启自动混合精度训练
- 采用ema策略(config.ema)
- 采用梯度积攒到一定数量再进行反向更新梯度策略(config.nominal_batch_factor)
请参考[DeepVAC](https://github.com/DeepVAC/deepvac)