當前位置：首頁 > 人工智能 > pytorch >内容正文

pytorch

【API调用】人脸检测+人脸属性（旷视 / 百度）

發布時間：2024/1/8 pytorch 33 豆豆

生活随笔收集整理的這篇文章主要介紹了【API调用】人脸检测+人脸属性（旷视 / 百度）小編覺得挺不錯的,現在分享給大家,幫大家做個參考.

文章目錄

1 簡介
2 曠視 face++
- 2.1 注冊
- - [獲得 API Key](https://console.faceplusplus.com.cn/app/apikey/list)
- 2.2 調用API(以人臉檢測為例)
- - 調用代碼
  - 將結果轉化為json
- 2.3 API返回結果
- - keys
  - 結果demo
- 2.3 異常
- - 連接超時
3. 百度
- 3.0 調用參考
- 3.1 [創建應用](https://console.bce.baidu.com/ai/?_=1629876388503&fromai=1#/ai/face/app/list)
- - 獲得apikey / secrect-key
- 3.2 調用代碼
- - 代碼需補充自己的信息
- 3.3 結果

1 簡介

本文，基于曠視與百度API，調用獲得人臉檢測以及人臉屬性信息（年齡性別情緒種族等），
大批量調用時，本人實驗發現（約調用·3萬次）
曠視API的調用代碼簡單，耗時短（約100ms一次）
以上結論，因本人水平有限，僅限于本人實驗。

2 曠視 face++

官網web-demo在線演示

2.1 注冊

獲得 API Key

登錄曠視人工智能開放平臺，注冊后，在平臺內申請API Key

2.2 調用API(以人臉檢測為例)

人臉檢測-文檔
人臉檢測-參考代碼
調用速度：約100毫秒/次

調用代碼

需要填上以上信息才能用

key = “填上你的KEY”
secret = “填上你的SECRET”
filepath = r"本地圖片的路徑"

# -*- coding: utf-8 -*- import urllib.request import urllib.error import timehttp_url = 'https://api-cn.faceplusplus.com/facepp/v3/detect' key = "填上你的KEY" secret = "填上你的SECRET" filepath= r"本地圖片的路徑"boundary = '----------%s' % hex(int(time.time() * 1000)) data = [] data.append('--%s' % boundary) data.append('Content-Disposition: form-data; name="%s"\r\n' % 'api_key') data.append(key) data.append('--%s' % boundary) data.append('Content-Disposition: form-data; name="%s"\r\n' % 'api_secret') data.append(secret) data.append('--%s' % boundary) fr = open(filepath, 'rb') data.append('Content-Disposition: form-data; name="%s"; filename=" "' % 'image_file') data.append('Content-Type: %s\r\n' % 'application/octet-stream') data.append(fr.read()) fr.close() data.append('--%s' % boundary) data.append('Content-Disposition: form-data; name="%s"\r\n' % 'return_landmark') data.append('1') data.append('--%s' % boundary) data.append('Content-Disposition: form-data; name="%s"\r\n' % 'return_attributes') data.append("gender,age,smiling,headpose,facequality,blur,eyestatus,emotion,ethnicity,beauty,mouthstatus,eyegaze,skinstatus") data.append('--%s--\r\n' % boundary)for i, d in enumerate(data):if isinstance(d, str):data[i] = d.encode('utf-8')http_body = b'\r\n'.join(data)# build http request req = urllib.request.Request(url=http_url, data=http_body)# header req.add_header('Content-Type', 'multipart/form-data; boundary=%s' % boundary)try:# post data to serverresp = urllib.request.urlopen(req, timeout=5)# get responseqrcont = resp.read()# if you want to load as json, you should decode first,# for example: json.loads(qrount.decode('utf-8'))print(qrcont.decode('utf-8')) except urllib.error.HTTPError as e:print(e.read().decode('utf-8'))

將結果轉化為json

修改處

# results=qrcont.decode('utf-8')results = json.loads(qrcont)

2.3 API返回結果

keys

dict_keys(['request_id', 'time_used', 'faces', 'image_id', 'face_num'])

結果demo

{"image_id": "Dd2xUw9S/7yjr0oDHHSL/Q==", "request_id": "1470472868,dacf2ff1-ea45-4842-9c07-6e8418cea78b", "time_used": 752, "faces": [{"landmark": {"mouth_upper_lip_left_contour2": {"y": 185, "x": 146}, "contour_chin": {"y": 231, "x": 137}, .........省略關鍵點信息"right_eye_pupil": {"y": 146, "x": 205}, "mouth_upper_lip_bottom": {"y": 195, "x": 159}}, "attributes": {"gender": {"value": "Female"}, "age": {"value": 21}, "glass": {"value": "None"}, "headpose": {"yaw_angle": -26.625063, "pitch_angle": 12.921974, "roll_angle": 22.814377}, "smile": {"threshold": 30.1, "value": 2.566890001296997}}, "face_rectangle": {"width": 140, "top": 89, "left": 104, "height": 141}, "face_token": "ed319e807e039ae669a4d1af0922a0c8"}],"face_num":1 }

2.3 異常

連接超時

CONCURRENCY_LIMIT_EXCEEDED

以上錯誤是，官方對免費的用戶請求頻率有限制，估計訪問頻次 2-5次/秒

3. 百度

https://ai.baidu.com/tech/face/detect
調用性能：約20秒1次

3.0 調用參考

https://blog.csdn.net/cskywit/article/details/81540500

3.1 創建應用

獲得apikey / secrect-key

文檔（人臉檢測）

其中client_id 為官網獲取的AcessKey， client_secret 為官網獲取的Secrectkey

# encoding:utf-8 import requests # client_id 為官網獲取的AK， client_secret 為官網獲取的SK host = 'https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id=【官網獲取的AK】&client_secret=【官網獲取的SK】' response = requests.get(host) if response:print(response.json())

3.2 調用代碼

首先需要獲得 access-token，

見代碼def get_access_token(self):
后續不需要執行，注釋 access_token=baidu_faceDetect_api.get_access_token()

代碼需補充自己的信息

import requests import base64 import json import timeclass BaiduFaceDetec():def __init__(self):self.api_key='你的'self.secret_key = '你的'self.access_token='你的'self.headers = {'content-type': 'application/json'}# self.headers = {# "Content-Type": "application/json; charset=UTF-8"# }def get_access_token(self):# client_id 為官網獲取的AK， client_secret 為官網獲取的SKhost = f'https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id={self.api_key}&client_secret={self.secret_key}'response = requests.get(host)if response:# print(response.json())self.access_token=response.json()['access_token']print("finish get access_token")return self.access_token"""{'refresh_token': xxx'expires_in': 2592000, 'session_key': xxx'access_token': xxxxx'scope': 'public brain_all_scope vis-faceverify_faceverify_h5-face-liveness vis-faceverify_FACE_V3 vis-faceverify_idl_face_merge vis-faceverify_FACE_EFFECT vis-faceverify_face_beauty vis-faceverify_face_feature_sdk wise_adapt lebo_resource_base lightservice_public hetu_basic lightcms_map_poi kaidian_kaidian ApsMisTest_Test權限 vis-classify_flower lpq_開放 cop_helloScope ApsMis_fangdi_permission smartapp_snsapi_base smartapp_mapp_dev_manage iop_autocar oauth_tp_app smartapp_smart_game_openapi oauth_sessionkey smartapp_swanid_verify smartapp_opensource_openapi smartapp_opensource_recapi fake_face_detect_開放Scope vis-ocr_虛擬人物助理 idl-video_虛擬人物助理 smartapp_component smartapp_search_plugin avatar_video_test', 'session_secret': 'eef5ffe19217d25dff3c435ddda8468a'}"""def img_to_BASE64(self,path):with open(path,'rb') as f:base64_data = base64.b64encode(f.read())return base64_data# 原文鏈接：https://blog.csdn.net/cskywit/article/details/81540500def get_face_attri(self,path,method="BASE64"):'''人臉檢測與屬性分析'''request_url = "https://aip.baidubce.com/rest/2.0/face/v3/detect"t0 = time.time()img_BASE64 = self.img_to_BASE64(path)t1 = time.time()# print("Time cost on one image {} s".format(t1 - t0))# print("img_BASE64",img_BASE64)post_data = {"image": img_BASE64,"image_type":method ,# "face_field": "gender,age,beauty,emotion,expression,faceshape,","face_field": "quality,gender,age,beauty,gender,emotion,expression,eye_status,landmark,faceshape,","face_type": "LIVE"}request_url = request_url + "?access_token=" + self.access_tokenresponse = requests.post(url=request_url, data=post_data, headers=self.headers)print(response.text)json_result = json.loads(response.text)print("detect_result:",json_result)if json_result['error_msg'] != 'pic not has face':print("圖片中包含人物年齡：", json_result['result']['face_list'][0]['age'])print("圖片中包含人物顏值評分：", json_result['result']['face_list'][0]['beauty'])print("圖片中包含人物性別：", json_result['result']['face_list'][0]['gender']['type'])# print("圖像品質：", json_result['result']['face_list'][0]['quality'])# 人臉旋轉角度參數# yaw 三維旋轉之左右旋轉角[-90(左), 90(右)]# pitch 三維旋轉之俯仰角度[-90(上), 90(下)]# roll 平面內旋轉角[-180(逆時針), 180(順時針)]print("人臉旋轉角度參數：", json_result['result']['face_list'][0]['angle'])print("是否微笑：", json_result['result']['face_list'][0]['expression']) # 笑不笑# angry:憤怒 disgust:厭惡 fear:恐懼 happy:高興 sad:傷心 surprise:驚訝 neutral:無表情 pouty: 撅嘴 grimace:鬼臉print("情緒如何", json_result['result']['face_list'][0]['emotion'])#左眼狀態 [0,1]取值，越接近0閉合的可能性越大print("是否睜眼", json_result['result']['face_list'][0]['eye_status'])# Press the green button in the gutter to run the script. if __name__ == '__main__':t0=time.time()img_path=r'你的本地圖片地址'baidu_faceDetect_api=BaiduFaceDetec()access_token=baidu_faceDetect_api.get_access_token()# print(access_token)baidu_faceDetect_api.get_face_attri(img_path)t1=time.time()print("Time cost on one image {} s".format(t1-t0))

3.3 結果

約20秒一次（也可能是我代碼有問題）

總結

以上是生活随笔為你收集整理的【API调用】人脸检测+人脸属性（旷视 / 百度）的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。

上一篇：紫光展锐发布系统级安全的高性能5G So
下一篇：学栈和队列时的人生感悟