AffectNet Preprocessing implementiert. Kleine Refactorings

projectmood/datasets/affectnet.py

+import argparse
+import logging
+import os
+import shutil
+import sys
+from shutil import copyfile
+
+import pandas as pd
+
+logfile = '../logs/affectnet.log'
+
+# Erstellt und konfiguriert die Logdatei
+logging.basicConfig(level=logging.NOTSET, format='%(asctime)s %(levelname)-8s %(message)s',
+                    datefmt='%m-%d %H:%M',
+                    filename=logfile)
+
+parser = argparse.ArgumentParser(description='AffectNet PrePipeline')
+parser.add_argument('csv', action='store', help='the .csv file to process')
+parser.add_argument('source', action='store', help='image source folder')
+parser.add_argument('-0', action='append_const', dest='emotions', const='neutral', help='neutral')
+parser.add_argument('-1', action='append_const', dest='emotions', const='happy', help='happy')
+parser.add_argument('-2', action='append_const', dest='emotions', const='sadness', help='sadness')
+parser.add_argument('-3', action='append_const', dest='emotions', const='surprise', help='surprise')
+parser.add_argument('-4', action='append_const', dest='emotions', const='fear', help='fear')
+parser.add_argument('-5', action='append_const', dest='emotions', const='disgust', help='disgust')
+parser.add_argument('-6', action='append_const', dest='emotions', const='anger', help='anger')
+parser.add_argument('-7', action='append_const', dest='emotions', const='contempt', help='contempt')
+parser.add_argument('-8', action='append_const', dest='emotions', const='none', help='none')
+parser.add_argument('-9', action='append_const', dest='emotions', const='uncertain', help='uncertain')
+parser.add_argument('-d', action='store', dest='destination', default='sorted_set', help='destination to sorted set')
+parser.add_argument('-l', '--limit', action='store', type=int, dest='limit', default=-1, help='image limit per emotion')
+parser.add_argument('-v', '--valence', action='store', default='-2', dest='valence', type=float, help='min valence')
+parser.add_argument('-a', '--arousal', action='store', default='-2', dest='arousal', type=float, help='min arousal')
+arguments = parser.parse_args()
+logging.debug(arguments)
+
+emodict = {0: 'neutral', 1: 'happy', 2: 'sadness', 3: 'surprise', 4: 'fear', 5: 'disgust', 6: 'anger', 7: 'contempt', 8: 'none', 9: 'uncertain', 10: 'noface'}
+
+counter = {}
+for e in arguments.emotions:
+    counter[e] = 0
+
+curdir = os.path.abspath(os.path.dirname(__file__))
+csv = os.path.join(curdir, arguments.csv)
+source = os.path.join(curdir, arguments.source)
+destination = os.path.join(curdir, arguments.destination)
+
+if os.path.exists(destination):
+    delete = input(destination + ' existiert. Durch diesen Prozess werden die Daten gelöscht. Fortfahren (y/n): ')
+    if delete == 'y':
+        shutil.rmtree(destination)
+    else:
+        logging.info('process aborted')
+        sys.exit()
+
+def generate_records(csvfile):
+    data = pd.read_csv(csvfile, delimiter=',', dtype='a')
+
+    imagepaths = pd.np.array(data['subDirectory_filePath'], pd.np.str)
+    labels = pd.np.array(data['expression'], pd.np.int)
+    valence = pd.np.array(data['valence'], pd.np.float)
+    arousal = pd.np.array(data['arousal'], pd.np.float)
+
+    for e in arguments.emotions:
+        dest = os.path.join(destination, e)
+        if not os.path.exists(dest):
+            os.makedirs(dest)
+
+    data = zip(labels, imagepaths, valence, arousal)
+
+    for d in data:
+        img = d[1]
+        pathfrom = os.path.join(source, img)
+
+        if not os.path.exists(pathfrom):
+            logging.info(img + ' not found')
+            continue
+
+        if all(count == arguments.limit for count in counter.values()):
+            break
+
+        emotion = emodict.get(d[0])
+
+        if emotion not in arguments.emotions or d[2] < float(arguments.valence) or d[3] < float(arguments.arousal) or counter[emotion] == arguments.limit:
+            continue
+
+        parent = os.path.join(destination, emotion)
+
+        pathto = os.path.join(parent, img.split('/')[1])
+        copyfile(pathfrom, pathto)
+        logging.info('wrote file ' + pathto)
+        counter[emotion] += 1
+
+
+if __name__ == '__main__':
+    generate_records(csv)
+
+    logging.info('finish')
+    logging.info(counter)

projectmood/image_sort.py → projectmood/datasets/cohn_kanade.py

@@ -14,7 +14,7 @@ emotions = ["neutral", "anger", "contempt", "disgust", "fear", "happy", "sadness
 """
 Liste mit den Dateiordnern aller Teilnehmer 
 """
-participants = glob.glob("resources/img_data/source_emotion/*")
+participants = glob.glob("../resources/img_data/source_emotion/*")
 
 for x in participants:
 
@@ -39,7 +39,7 @@ for x in participants:
             """
             emotion = int(float(file.readline()))
 
-            source_emotions = glob.glob("resources/img_data/source_images/%s/%s/*.png" % (number, current_session))
+            source_emotions = glob.glob("../resources/img_data/source_images/%s/%s/*.png" % (number, current_session))
             source_emotions.sort()
 
             """
@@ -58,8 +58,8 @@ for x in participants:
             Für den neutralen Ausdruck 
             Für die Emotion 
             """
-            dest_neut = "resources/img_data/sorted_set/neutral/%s" % sourcefile_neutral[44:]
-            dest_emot = "resources/img_data/sorted_set/%s/%s" % (emotions[emotion], sourcefile_emotion[44:])
+            dest_neut = "../resources/img_data/sorted_set/neutral/%s" % sourcefile_neutral[44:]
+            dest_emot = "../resources/img_data/sorted_set/%s/%s" % (emotions[emotion], sourcefile_emotion[44:])
 
             """Kopiert Dateien"""
             copyfile(sourcefile_neutral, dest_neut)

projectmood/datasets/fer2013.py

+import numpy as np
+import cv2
+import mxnet as mx
+import pandas as pd
+import random
+import os
+
+curdir = os.path.abspath(os.path.dirname(__file__))
+
+def gen_record(csvfile,channel):
+    data = pd.read_csv(csvfile,delimiter=',',dtype='a')
+    labels = np.array(data['emotion'],np.float)
+    # print(labels,'\n',data['emotion'])
+        
+    imagebuffer = np.array(data['pixels'])
+    images = np.array([np.fromstring(image,np.uint8,sep=' ') for image in imagebuffer])
+    del imagebuffer
+    num_shape = int(np.sqrt(images.shape[-1]))
+    images.shape = (images.shape[0],num_shape,num_shape)
+    # img=images[0];cv2.imshow('test',img);cv2.waitKey(0);cv2.destroyAllWindow();exit()
+    dirs = set(data['Usage'])
+    subdirs = set(labels)
+    class_dir = {}
+    for dr in dirs:
+        dest = os.path.join(curdir,dr)
+        class_dir[dr] = dest
+        if not os.path.exists(dest):
+            os.mkdir(dest)
+            
+    data = zip(labels,images,data['Usage'])
+    
+    for d in data:
+        destdir = os.path.join(class_dir[d[-1]],str(int(d[0])))
+        if not os.path.exists(destdir):
+            os.mkdir(destdir)
+        img = d[1]
+        filepath = unique_name(destdir,d[-1])
+        print('[^_^] Write image to %s' % filepath)
+        if not filepath:
+            continue
+        sig = cv2.imwrite(filepath,img)
+        if not sig:
+            print('Error')
+            exit(-1)
+
+
+def unique_name(pardir,prefix,suffix='jpg'):
+    filename = '{0}_{1}.{2}'.format(prefix,random.randint(1,10**8),suffix)
+    filepath = os.path.join(pardir,filename)
+    if not os.path.exists(filepath):
+        return filepath
+    unique_name(pardir,prefix,suffix)
+    
+
+
+if __name__ == '__main__':
+    filename = 'fer2013.csv'
+    filename = os.path.join(curdir,filename)
+    gen_record(filename,1)
+    
+    # ##################### test
+    # tmp = unique_name('./Training','Training')
+    # print(tmp)
\ No newline at end of file