架構
1.web服務用tornado；
2.模型訓練和預測用tensorflow；
3.opencv做人臉檢測
Prerequisites:
- tensorflow0.12
- opencv3.0
- python2.7
- 訓練好的模型，模型訓練參考https://github.com/dpressel/rude-carnie
主代碼
# code: UTF-8
import tornado.ioloop
import tornado.web
import uuid
#import Image
import StringIO
import os
import commands
import guess
import json
class PredictHandler(tornado.web.RequestHandler):
def get(self):
self.render("index.html")
def post(self):
if self.request.files:
file_name = "%s" % uuid.uuid1()
print 'file_name',file_name
file_raw = self.request.files["file"][0]["body"]
usr_home = os.path.expanduser('~')
file_name = usr_home+"/tensorflow/static/tmp/m_%s.jpg" % file_name
fin = open(file_name,"w")
print "success to open file"
fin.write(file_raw)
fin.close()
print "use tensorflow"
#首先檢測是否是人類，如果不是，直接返回
isman = guess.detectface(file_name)
feeds_json = {}
if isman == 1:
#預測年齡、顏值打分
age,score = guess.guessAge(file_name)
#預測性別
gender = guess.guessGender(file_name)
print 'guess age is ', age
print 'gender is ',gender
dic = {}
dic['isman'] = True
dic['age'] = age
dic['score'] = score
dic['gender'] = gender
feeds_json = json.dumps(dic)
elif isman == 0 :
dic = {}
dic['isman'] = False
feeds_json = json.dumps(dic)
self.set_header('Content-Type', 'application/json; charset=UTF-8')
self.write(feeds_json)
self.finish()
application = tornado.web.Application([
(r"/predict", PredictHandler),
])
if __name__ == "__main__":
application.listen(8882)
tornado.ioloop.IOLoop.instance().start()使用到了guess.py，代碼如下
#code: UTF-8
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import sys
reload(sys)
sys.setdefaultencoding('utf-8')
from datetime import datetime
import math
import time
from data import inputs
import numpy as np
import tensorflow as tf
from model import select_model, get_checkpoint
from utils import ImageCoder, make_batch, FaceDetector
import os
import csv
import random
RESIZE_FINAL = 227
GENDER_LIST =['M','F']
AGE_LIST = ['(0, 2)','(4, 6)','(8, 12)','(15, 20)','(25, 32)','(38, 43)','(48, 53)','(60, 100)']
AGE_MODEL_PATH = '/work/facescore/22801'
GENDER_MODEL_PATH = '/run-1322'
model_checkpoint_path = ''
tf.app.flags.DEFINE_string('model_dir', '',
'Model directory (where training data lives)')
tf.app.flags.DEFINE_string('class_type', 'age',
'Classification type (age|gender)')
tf.app.flags.DEFINE_string('device_id', '/cpu:0',
'What processing unit to execute inference on')
tf.app.flags.DEFINE_string('filename', '',
'File (Image) or File list (Text/No header TSV) to process')
tf.app.flags.DEFINE_string('target', '',
'CSV file containing the filename processed along with best guess and score')
tf.app.flags.DEFINE_string('checkpoint', 'checkpoint',
'Checkpoint basename')
tf.app.flags.DEFINE_string('model_type', 'default',
'Type of convnet')
tf.app.flags.DEFINE_string('requested_step', '', 'Within the model directory, a requested step to restore e.g., 9000')
tf.app.flags.DEFINE_boolean('single_look', False, 'single look at the image or multiple crops')
tf.app.flags.DEFINE_string('face_detection_model', '', 'Do frontal face detection with model specified')
FLAGS = tf.app.flags.FLAGS
def one_of(fname, types):
for ty in types:
if fname.endswith('.' + ty):
return True
return False
def resolve_file(fname):
if os.path.exists(fname): return fname
for suffix in ('.jpg', '.png', '.JPG', '.PNG', '.jpeg'):
cand = fname + suffix
if os.path.exists(cand):
return cand
return None
def classify(sess, label_list, softmax_output, coder, images, image_file):
print('Running file %s' % image_file)
image_batch = make_batch(image_file, coder, not FLAGS.single_look)
batch_results = sess.run(softmax_output, feed_dict={images:image_batch.eval()})
output = batch_results[0]
batch_sz = batch_results.shape[0]
for i in range(1, batch_sz):
output = output + batch_results[i]
output /= batch_sz
best = np.argmax(output)
best_choice = (label_list[best], output[best])
print('Guess @ 1 %s, prob = %.2f' % best_choice)
#calculate face score
score = scoreAge(output)
nlabels = len(label_list)
if nlabels > 2:
output[best] = 0
second_best = np.argmax(output)
print('Guess @ 2 %s, prob = %.2f' % (label_list[second_best], output[second_best]))
return label_list[best],score
def classifyGender(sess, label_list, softmax_output, coder, images, image_file):
print('Running file %s' % image_file)
image_batch = make_batch(image_file, coder, not FLAGS.single_look)
batch_results = sess.run(softmax_output, feed_dict={images:image_batch.eval()})
output = batch_results[0]
batch_sz = batch_results.shape[0]
for i in range(1, batch_sz):
output = output + batch_results[i]
output /= batch_sz
best = np.argmax(output)
best_choice = (label_list[best], output[best])
print('Guess @ 1 %s, prob = %.2f' % best_choice)
#calculate face score
#score = scoreAge(output)
nlabels = len(label_list)
if nlabels > 2:
output[best] = 0
second_best = np.argmax(output)
print('Guess @ 2 %s, prob = %.2f' % (label_list[second_best], output[second_best]))
return label_list[best]
def scoreAge(prop_list):
#['(0, 2)', '(4, 6)', '(8, 12)', '(15, 20)', '(25, 32)', '(38, 43)', '(48, 53)', '(60, 100)']
#[90, 80, 80, 70, 60, 50, 40, 20]
score_list = [100, 100, 95, 90, 80, 50, 40, 20]
random_score_list = []
finalScore = 0
for score in score_list:
randScore = score + random.random() * 10
random_score_list.append(randScore)
j = 0
while(j < len(random_score_list)):
prop = prop_list[j]
score = random_score_list[j]
finalScore = finalScore + score * prop
j = j + 1
return finalScore
def batchlist(srcfile):
with open(srcfile, 'r') as csvfile:
reader = csv.reader(csvfile)
if srcfile.endswith('.csv') or srcfile.endswith('.tsv'):
print('skipping header')
reader.next()
return [row[0] for row in reader]
def detectface(filename):
files = []
#print('Using face detector %s' % FLAGS.face_detection_model)
face_detect = FaceDetector('haarcascade_frontalface_default.xml')
face_files, rectangles = face_detect.run(filename)
files += face_files
if (len(files)>0) :
return 1
else:
return 0
def guessAge(image_file):
#import!!!Fix the bug http://stackoverflow.com/questions/33765336/remove-nodes-from-graph-or-reset-entire-default-graph
tf.reset_default_graph()
with tf.Session() as sess:
age_label_list = AGE_LIST
agelabels = len(age_label_list)
# print('Executing on %s' % FLAGS.device_id)
model_fn = select_model('inception')
images = tf.placeholder(tf.float32, [None, RESIZE_FINAL, RESIZE_FINAL, 3])
logits_age = model_fn(agelabels, images, 1, False)
init = tf.initialize_all_variables()
requested_step = FLAGS.requested_step if FLAGS.requested_step else None
checkpoint_path = '%s' % (AGE_MODEL_PATH)
# update in 0.11 version
model_checkpoint_path, global_step = get_checkpoint("/22801/", requested_step, FLAGS.checkpoint)
#print 'model_checkpoint_path is', model_checkpoint_path
#print model_checkpoint_path
saver = tf.train.Saver()
if not saver.last_checkpoints :
saver.restore(sess, model_checkpoint_path)
softmax_output = tf.nn.softmax(logits_age)
coder = ImageCoder()
files = []
# detect age
best_choice = classify(sess, age_label_list, softmax_output, coder, images, image_file)
sess.close()
return best_choice
def guessGender(image_file):
tf.reset_default_graph()
with tf.Session() as sess:
sess = tf.Session()
age_label_list = AGE_LIST
gender_label_list = GENDER_LIST
genderlabels = len(gender_label_list)
# print('Executing on %s' % FLAGS.device_id)
model_fn = select_model('')
images = tf.placeholder(tf.float32, [None, RESIZE_FINAL, RESIZE_FINAL, 3])
logits_gender = model_fn(genderlabels, images, 1, False)
init = tf.initialize_all_variables()
requested_step = 10000
checkpoint_path = '%s' % (GENDER_MODEL_PATH)
model_checkpoint_path, global_step = get_checkpoint(checkpoint_path, requested_step, FLAGS.checkpoint)
saver = tf.train.Saver()
saver.restore(sess, model_checkpoint_path)
softmax_output = tf.nn.softmax(logits_gender)
coder = ImageCoder()
files = []
# detect gender
#try:
best_choice = classifyGender(sess, gender_label_list, softmax_output, coder, images, image_file)
return best_choice
#except Exception as e:
# print(e)
# print('Failed to run image %s ' % image_file)