paddlehub -- 开箱即用的模型库

PaddleHub

https://github.com/PaddlePaddle/PaddleHub

令人惊叹的已训练好的模型工具库， 基于Paddle。

Awesome pre-trained models toolkit based on PaddlePaddle.（260+ models including Image, Text, Audio and Video with Easy Inference & Serving deployment)

提供丰富、高质量、直接可用的已训练好的模型

不需要深度学习背景

覆盖四大类别，图像、文本、音频、视频

开源、免费

Introduction

• PaddleHub aims to provide developers with rich, high-quality, and directly usable pre-trained models.
• No need for deep learning background, you can use AI models quickly and enjoy the dividends of the artificial intelligence era.
• Covers 4 major categories of Image, Text, Audio, and Video, and supports one-click prediction, easy service deployment and transfer learning
• All models are OPEN SOURCE, FREE to download and use them in offline scenario.

 

特定模型服务于特定场景

https://www.paddlepaddle.org.cn/hub

PaddleHub

便捷地获取PaddlePaddle生态下的预训练模型，完成模型的管理和一键预测。配合使用Fine-tune API，可以基于大规模预训练模型快速完成迁移学习，让预训练模型能更好地服务于用户特定场景的应用。

• 无需数据和训练，一键模型应用
• 一键模型转服务
• 易用的迁移学习
• 丰富的预训练模型

安装两个库

!pip install --upgrade paddlepaddle -i https://mirror.baidu.com/pypi/simple
!pip install --upgrade paddlehub -i https://mirror.baidu.com/pypi/simple

 

示例

几行代码就可使用。

如下是中文分词工具使用

!pip install --upgrade paddlepaddle -i https://mirror.baidu.com/pypi/simple
!pip install --upgrade paddlehub -i https://mirror.baidu.com/pypi/simple

import paddlehub as hub

lac = hub.Module(name="lac")
test_text = ["今天是个好天气。"]

results = lac.cut(text=test_text, use_gpu=False, batch_size=1, return_tag=True)
print(results)
#{'word': ['今天', '是', '个', '好天气', '。'], 'tag': ['TIME', 'v', 'q', 'n', 'w']}

 

模型库-modelbase

https://www.paddlepaddle.org.cn/modelbase

• 智能视觉(PaddleCV)
  • 图像分类
  • 目标检测
  • 图像分割
  • 关键点检测
  • 图像生成
  • 场景文字识别
  • 度量学习
  • 视频
• 智能文本处理(PaddleNLP)
  • NLP 基础技术
  • NLP 核心技术
  • NLP系统应用
• 智能推荐(PaddleRec)
• 智能语音(PaddleSpeech)
• 其他模型

 

 

chinese_ocr_db_crnn_mobile

https://www.paddlepaddle.org.cn/hubdetail?name=chinese_ocr_db_crnn_mobile&en_category=TextRecognition

 

支持中文的OCR模型。

支持三种使用方式

命令行预测

$ hub run chinese_ocr_db_crnn_mobile --input_path "/PATH/TO/IMAGE"

API调用

import paddlehub as hub
import cv2

ocr = hub.Module(name="chinese_ocr_db_crnn_mobile")
result = ocr.recognize_text(images=[cv2.imread('/PATH/TO/IMAGE')])

# or
# result = ocr.recognize_text(paths=['/PATH/TO/IMAGE'])

服务部署

启动PaddleHub Serving

运行启动命令：

$ hub serving start -m chinese_ocr_db_crnn_mobile

发送预测请求

配置好服务端，以下数行代码即可实现发送预测请求，获取预测结果

import requests
import json
import cv2
import base64

def cv2_to_base64(image):
    data = cv2.imencode('.jpg', image)[1]
    return base64.b64encode(data.tostring()).decode('utf8')

# 发送HTTP请求
data = {'images':[cv2_to_base64(cv2.imread("/PATH/TO/IMAGE"))]}
headers = {"Content-type": "application/json"}
url = "http://127.0.0.1:8866/predict/chinese_ocr_db_crnn_mobile"
r = requests.post(url=url, headers=headers, data=json.dumps(data))

# 打印预测结果
print(r.json()["results"])

 

DEMO

https://github.com/fanqingsong/code_snippet/blob/master/machine_learning/paddle/ocr.py

 

从验证码中提取数字

 

 

import paddlehub as hub
import cv2

ocr = hub.Module(name="chinese_ocr_db_crnn_mobile")
result = ocr.recognize_text(images=[cv2.imread('./test2.png')])

print(result)

 

如下为打印，粗体为提取数字。

 WARNING: Logging before InitGoogleLogging() is written to STDERR
W0310 14:03:05.659210 11502 default_variables.cpp:429] Fail to open /proc/self/io: No such file or directory [2]
/root/.pyenv/versions/3.6.8/lib/python3.6/site-packages/setuptools/depends.py:2: DeprecationWarning: the imp module is deprecated in favour of importlib; see the module's documentation for alternative uses
  import imp
[2021-03-10 14:03:14,697] [ WARNING] - The _initialize method in HubModule will soon be deprecated, you can use the __init__() to handle the initialization of the object
W0310 14:03:14.708413 11502 analysis_predictor.cc:1145] Deprecated. Please use CreatePredictor instead.
[2021-03-10 14:03:15,063] [ WARNING] - The _initialize method in HubModule will soon be deprecated, you can use the __init__() to handle the initialization of the object
[{'save_path': '', 'data': [{'text': '6067', 'confidence': 0.8805994987487793, 'text_box_position': [[9, 2], [52, 2], [52, 16], [9, 16]]}]}]